From 7915a15b2c7d8e31422c472654b078e2ef6b50e6 Mon Sep 17 00:00:00 2001 From: haoyuying <35907364+haoyuying@users.noreply.github.com> Date: Wed, 21 Oct 2020 15:15:53 +0800 Subject: [PATCH] add efficientnet b1 ~ b7 --- .../efficientnetb1_imagenet/README.md | 149 --- .../efficientnetb1_imagenet/__init__.py | 1 - .../efficientnetb1_imagenet/data_feed.py | 98 -- .../efficientnetb1_imagenet/efficientnet.py | 623 ---------- .../efficientnetb1_imagenet/label_list.txt | 1000 ---------------- .../efficientnetb1_imagenet/layers.py | 248 ---- .../efficientnetb1_imagenet/module.py | 1049 ++++++++++++----- .../efficientnetb1_imagenet/processor.py | 69 -- .../efficientnetb2_imagenet/README.md | 149 --- .../efficientnetb2_imagenet/__init__.py | 1 - .../efficientnetb2_imagenet/data_feed.py | 98 -- .../efficientnetb2_imagenet/efficientnet.py | 623 ---------- .../efficientnetb2_imagenet/label_list.txt | 1000 ---------------- .../efficientnetb2_imagenet/layers.py | 248 ---- .../efficientnetb2_imagenet/module.py | 1046 +++++++++++----- .../efficientnetb2_imagenet/processor.py | 69 -- .../efficientnetb3_imagenet/README.md | 149 --- .../efficientnetb3_imagenet/__init__.py | 1 - .../efficientnetb3_imagenet/data_feed.py | 98 -- .../efficientnetb3_imagenet/efficientnet.py | 623 ---------- .../efficientnetb3_imagenet/label_list.txt | 1000 ---------------- .../efficientnetb3_imagenet/layers.py | 248 ---- .../efficientnetb3_imagenet/module.py | 1046 +++++++++++----- .../efficientnetb3_imagenet/processor.py | 69 -- .../efficientnetb4_imagenet/README.md | 149 --- .../efficientnetb4_imagenet/__init__.py | 1 - .../efficientnetb4_imagenet/data_feed.py | 98 -- .../efficientnetb4_imagenet/efficientnet.py | 623 ---------- .../efficientnetb4_imagenet/label_list.txt | 1000 ---------------- .../efficientnetb4_imagenet/layers.py | 248 ---- .../efficientnetb4_imagenet/module.py | 1046 +++++++++++----- .../efficientnetb4_imagenet/processor.py | 69 -- .../efficientnetb5_imagenet/README.md | 149 --- .../efficientnetb5_imagenet/__init__.py | 1 - .../efficientnetb5_imagenet/data_feed.py | 98 -- .../efficientnetb5_imagenet/efficientnet.py | 623 ---------- .../efficientnetb5_imagenet/label_list.txt | 1000 ---------------- .../efficientnetb5_imagenet/layers.py | 248 ---- .../efficientnetb5_imagenet/module.py | 1040 +++++++++++----- .../efficientnetb5_imagenet/processor.py | 69 -- .../efficientnetb6_imagenet/README.md | 149 --- .../efficientnetb6_imagenet/__init__.py | 1 - .../efficientnetb6_imagenet/data_feed.py | 98 -- .../efficientnetb6_imagenet/efficientnet.py | 623 ---------- .../efficientnetb6_imagenet/label_list.txt | 1000 ---------------- .../efficientnetb6_imagenet/layers.py | 248 ---- .../efficientnetb6_imagenet/module.py | 1040 +++++++++++----- .../efficientnetb6_imagenet/processor.py | 69 -- .../efficientnetb7_imagenet/README.md | 149 --- .../efficientnetb7_imagenet/__init__.py | 1 - .../efficientnetb7_imagenet/data_feed.py | 98 -- .../efficientnetb7_imagenet/efficientnet.py | 623 ---------- .../efficientnetb7_imagenet/label_list.txt | 1000 ---------------- .../efficientnetb7_imagenet/layers.py | 248 ---- .../efficientnetb7_imagenet/module.py | 1040 +++++++++++----- .../efficientnetb7_imagenet/processor.py | 69 -- 56 files changed, 5159 insertions(+), 17464 deletions(-) delete mode 100644 hub_module/modules/image/classification/efficientnetb1_imagenet/README.md delete mode 100644 hub_module/modules/image/classification/efficientnetb1_imagenet/__init__.py delete mode 100644 hub_module/modules/image/classification/efficientnetb1_imagenet/data_feed.py delete mode 100644 hub_module/modules/image/classification/efficientnetb1_imagenet/efficientnet.py delete mode 100644 hub_module/modules/image/classification/efficientnetb1_imagenet/label_list.txt delete mode 100644 hub_module/modules/image/classification/efficientnetb1_imagenet/layers.py delete mode 100644 hub_module/modules/image/classification/efficientnetb1_imagenet/processor.py delete mode 100644 hub_module/modules/image/classification/efficientnetb2_imagenet/README.md delete mode 100644 hub_module/modules/image/classification/efficientnetb2_imagenet/__init__.py delete mode 100644 hub_module/modules/image/classification/efficientnetb2_imagenet/data_feed.py delete mode 100644 hub_module/modules/image/classification/efficientnetb2_imagenet/efficientnet.py delete mode 100644 hub_module/modules/image/classification/efficientnetb2_imagenet/label_list.txt delete mode 100644 hub_module/modules/image/classification/efficientnetb2_imagenet/layers.py delete mode 100644 hub_module/modules/image/classification/efficientnetb2_imagenet/processor.py delete mode 100644 hub_module/modules/image/classification/efficientnetb3_imagenet/README.md delete mode 100644 hub_module/modules/image/classification/efficientnetb3_imagenet/__init__.py delete mode 100644 hub_module/modules/image/classification/efficientnetb3_imagenet/data_feed.py delete mode 100644 hub_module/modules/image/classification/efficientnetb3_imagenet/efficientnet.py delete mode 100644 hub_module/modules/image/classification/efficientnetb3_imagenet/label_list.txt delete mode 100644 hub_module/modules/image/classification/efficientnetb3_imagenet/layers.py delete mode 100644 hub_module/modules/image/classification/efficientnetb3_imagenet/processor.py delete mode 100644 hub_module/modules/image/classification/efficientnetb4_imagenet/README.md delete mode 100644 hub_module/modules/image/classification/efficientnetb4_imagenet/__init__.py delete mode 100644 hub_module/modules/image/classification/efficientnetb4_imagenet/data_feed.py delete mode 100644 hub_module/modules/image/classification/efficientnetb4_imagenet/efficientnet.py delete mode 100644 hub_module/modules/image/classification/efficientnetb4_imagenet/label_list.txt delete mode 100644 hub_module/modules/image/classification/efficientnetb4_imagenet/layers.py delete mode 100644 hub_module/modules/image/classification/efficientnetb4_imagenet/processor.py delete mode 100644 hub_module/modules/image/classification/efficientnetb5_imagenet/README.md delete mode 100644 hub_module/modules/image/classification/efficientnetb5_imagenet/__init__.py delete mode 100644 hub_module/modules/image/classification/efficientnetb5_imagenet/data_feed.py delete mode 100644 hub_module/modules/image/classification/efficientnetb5_imagenet/efficientnet.py delete mode 100644 hub_module/modules/image/classification/efficientnetb5_imagenet/label_list.txt delete mode 100644 hub_module/modules/image/classification/efficientnetb5_imagenet/layers.py delete mode 100644 hub_module/modules/image/classification/efficientnetb5_imagenet/processor.py delete mode 100644 hub_module/modules/image/classification/efficientnetb6_imagenet/README.md delete mode 100644 hub_module/modules/image/classification/efficientnetb6_imagenet/__init__.py delete mode 100644 hub_module/modules/image/classification/efficientnetb6_imagenet/data_feed.py delete mode 100644 hub_module/modules/image/classification/efficientnetb6_imagenet/efficientnet.py delete mode 100644 hub_module/modules/image/classification/efficientnetb6_imagenet/label_list.txt delete mode 100644 hub_module/modules/image/classification/efficientnetb6_imagenet/layers.py delete mode 100644 hub_module/modules/image/classification/efficientnetb6_imagenet/processor.py delete mode 100644 hub_module/modules/image/classification/efficientnetb7_imagenet/README.md delete mode 100644 hub_module/modules/image/classification/efficientnetb7_imagenet/__init__.py delete mode 100644 hub_module/modules/image/classification/efficientnetb7_imagenet/data_feed.py delete mode 100644 hub_module/modules/image/classification/efficientnetb7_imagenet/efficientnet.py delete mode 100644 hub_module/modules/image/classification/efficientnetb7_imagenet/label_list.txt delete mode 100644 hub_module/modules/image/classification/efficientnetb7_imagenet/layers.py delete mode 100644 hub_module/modules/image/classification/efficientnetb7_imagenet/processor.py diff --git a/hub_module/modules/image/classification/efficientnetb1_imagenet/README.md b/hub_module/modules/image/classification/efficientnetb1_imagenet/README.md deleted file mode 100644 index cb0e1a4b..00000000 --- a/hub_module/modules/image/classification/efficientnetb1_imagenet/README.md +++ /dev/null @@ -1,149 +0,0 @@ -## 命令行预测 - -``` -hub run efficientnetb1_imagenet --input_path "/PATH/TO/IMAGE" -``` - -## API - -```python -def get_expected_image_width() -``` - -返回预处理的图片宽度，也就是224。 - -```python -def get_expected_image_height() -``` - -返回预处理的图片高度，也就是224。 - -```python -def get_pretrained_images_mean() -``` - -返回预处理的图片均值，也就是 \[0.485, 0.456, 0.406\]。 - -```python -def get_pretrained_images_std() -``` - -返回预处理的图片标准差，也就是 \[0.229, 0.224, 0.225\]。 - - -```python -def context(trainable=True, pretrained=True) -``` - -**参数** - -* trainable (bool): 计算图的参数是否为可训练的； -* pretrained (bool): 是否加载默认的预训练模型。 - -**返回** - -* inputs (dict): 计算图的输入，key 为 'image', value 为图片的张量； -* outputs (dict): 计算图的输出，key 为 'classification' 和 'feature_map'，其相应的值为： - * classification (paddle.fluid.framework.Variable): 分类结果，也就是全连接层的输出； - * feature\_map (paddle.fluid.framework.Variable): 特征匹配，全连接层前面的那个张量。 -* context\_prog(fluid.Program): 计算图，用于迁移学习。 - -```python -def classify(images=None, - paths=None, - batch_size=1, - use_gpu=False, - top_k=1): -``` - -**参数** - -* images (list\[numpy.ndarray\]): 图片数据，每一个图片数据的shape 均为 \[H, W, C\]，颜色空间为 BGR； -* paths (list\[str\]): 图片的路径； -* batch\_size (int): batch 的大小； -* use\_gpu (bool): 是否使用 GPU 来预测； -* top\_k (int): 返回预测结果的前 k 个。 - -**返回** - -res (list\[dict\]): 分类结果，列表的每一个元素均为字典，其中 key 为识别动物的类别，value为置信度。 - -```python -def save_inference_model(dirname, - model_filename=None, - params_filename=None, - combined=True) -``` - -将模型保存到指定路径。 - -**参数** - -* dirname: 存在模型的目录名称 -* model\_filename: 模型文件名称，默认为\_\_model\_\_ -* params\_filename: 参数文件名称，默认为\_\_params\_\_(仅当`combined`为True时生效) -* combined: 是否将参数保存到统一的一个文件中 - -## 代码示例 - -```python -import paddlehub as hub -import cv2 - -classifier = hub.Module(name="efficientnetb1_imagenet") - -result = classifier.classify(images=[cv2.imread('/PATH/TO/IMAGE')]) -# or -# result = classifier.classify(paths=['/PATH/TO/IMAGE']) -``` - -## 服务部署 - -PaddleHub Serving可以部署一个在线图像识别服务。 - -## 第一步：启动PaddleHub Serving - -运行启动命令： -```shell -$ hub serving start -m efficientnetb1_imagenet -``` - -这样就完成了一个在线图像识别服务化API的部署，默认端口号为8866。 - -**NOTE:** 如使用GPU预测，则需要在启动服务之前，请设置CUDA\_VISIBLE\_DEVICES环境变量，否则不用设置。 - -## 第二步：发送预测请求 - -配置好服务端，以下数行代码即可实现发送预测请求，获取预测结果 - -```python -import requests -import json -import cv2 -import base64 - - -def cv2_to_base64(image): - data = cv2.imencode('.jpg', image)[1] - return base64.b64encode(data.tostring()).decode('utf8') - - -# 发送HTTP请求 -data = {'images':[cv2_to_base64(cv2.imread("/PATH/TO/IMAGE"))]} -headers = {"Content-type": "application/json"} -url = "http://127.0.0.1:8866/predict/efficientnetb1_imagenet" -r = requests.post(url=url, headers=headers, data=json.dumps(data)) - -# 打印预测结果 -print(r.json()["results"]) -``` - -### 查看代码 - -https://github.com/PaddlePaddle/PaddleClas - -### 依赖 - -paddlepaddle >= 1.6.2 - -paddlehub >= 1.6.0 diff --git a/hub_module/modules/image/classification/efficientnetb1_imagenet/__init__.py b/hub_module/modules/image/classification/efficientnetb1_imagenet/__init__.py deleted file mode 100644 index 19a990da..00000000 --- a/hub_module/modules/image/classification/efficientnetb1_imagenet/__init__.py +++ /dev/null @@ -1 +0,0 @@ -# -*- coding:utf-8 -*- diff --git a/hub_module/modules/image/classification/efficientnetb1_imagenet/data_feed.py b/hub_module/modules/image/classification/efficientnetb1_imagenet/data_feed.py deleted file mode 100644 index 11181458..00000000 --- a/hub_module/modules/image/classification/efficientnetb1_imagenet/data_feed.py +++ /dev/null @@ -1,98 +0,0 @@ -# -*- coding:utf-8 -*- -# copyright (c) 2020 PaddlePaddle Authors. All Rights Reserve. -# -# Licensed under the Apache License, Version 2.0 (the "License"); -# you may not use this file except in compliance with the License. -# You may obtain a copy of the License at -# -# http://www.apache.org/licenses/LICENSE-2.0 -# -# Unless required by applicable law or agreed to in writing, software -# distributed under the License is distributed on an "AS IS" BASIS, -# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. -# See the License for the specific language governing permissions and -# limitations under the License. - -import os -import time -from collections import OrderedDict - -import cv2 -import numpy as np -from PIL import Image - -__all__ = ['reader'] - -DATA_DIM = 224 -img_mean = np.array([0.485, 0.456, 0.406]).reshape((3, 1, 1)) -img_std = np.array([0.229, 0.224, 0.225]).reshape((3, 1, 1)) - - -def resize_short(img, target_size): - percent = float(target_size) / min(img.size[0], img.size[1]) - resized_width = int(round(img.size[0] * percent)) - resized_height = int(round(img.size[1] * percent)) - img = img.resize((resized_width, resized_height), Image.LANCZOS) - return img - - -def crop_image(img, target_size, center): - width, height = img.size - size = target_size - if center == True: - w_start = (width - size) / 2 - h_start = (height - size) / 2 - else: - w_start = np.random.randint(0, width - size + 1) - h_start = np.random.randint(0, height - size + 1) - w_end = w_start + size - h_end = h_start + size - img = img.crop((w_start, h_start, w_end, h_end)) - return img - - -def process_image(img): - img = resize_short(img, target_size=256) - img = crop_image(img, target_size=DATA_DIM, center=True) - if img.mode != 'RGB': - img = img.convert('RGB') - img = np.array(img).astype('float32').transpose((2, 0, 1)) / 255 - img -= img_mean - img /= img_std - return img - - -def reader(images=None, paths=None): - """ - Preprocess to yield image. - - Args: - images (list[numpy.ndarray]): images data, shape of each is [H, W, C]. - paths (list[str]): paths to images. - - Yield: - each (collections.OrderedDict): info of original image, preprocessed image. - """ - component = list() - if paths: - for im_path in paths: - each = OrderedDict() - assert os.path.isfile( - im_path), "The {} isn't a valid file path.".format(im_path) - each['org_im_path'] = im_path - each['org_im'] = Image.open(im_path) - each['org_im_width'], each['org_im_height'] = each['org_im'].size - component.append(each) - if images is not None: - assert type(images), "images is a list." - for im in images: - each = OrderedDict() - each['org_im'] = Image.fromarray(im[:, :, ::-1]) - each['org_im_path'] = 'ndarray_time={}'.format( - round(time.time(), 6) * 1e6) - each['org_im_width'], each['org_im_height'] = each['org_im'].size - component.append(each) - - for element in component: - element['image'] = process_image(element['org_im']) - yield element diff --git a/hub_module/modules/image/classification/efficientnetb1_imagenet/efficientnet.py b/hub_module/modules/image/classification/efficientnetb1_imagenet/efficientnet.py deleted file mode 100644 index c357fa9c..00000000 --- a/hub_module/modules/image/classification/efficientnetb1_imagenet/efficientnet.py +++ /dev/null @@ -1,623 +0,0 @@ -# -*- coding:utf-8 -*- -# copyright (c) 2020 PaddlePaddle Authors. All Rights Reserve. -# -# Licensed under the Apache License, Version 2.0 (the "License"); -# you may not use this file except in compliance with the License. -# You may obtain a copy of the License at -# -# http://www.apache.org/licenses/LICENSE-2.0 -# -# Unless required by applicable law or agreed to in writing, software -# distributed under the License is distributed on an "AS IS" BASIS, -# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. -# See the License for the specific language governing permissions and -# limitations under the License. - -from __future__ import absolute_import -from __future__ import division -from __future__ import print_function - -import collections -import re -import math -import copy - -import paddle.fluid as fluid -from efficientnetb1_imagenet.layers import conv2d, init_batch_norm_layer, init_fc_layer - -__all__ = [ - 'EfficientNet', 'EfficientNetB0_small', 'EfficientNetB0', 'EfficientNetB1', - 'EfficientNetB2', 'EfficientNetB3', 'EfficientNetB4', 'EfficientNetB5', - 'EfficientNetB6', 'EfficientNetB7' -] - -GlobalParams = collections.namedtuple('GlobalParams', [ - 'batch_norm_momentum', - 'batch_norm_epsilon', - 'dropout_rate', - 'num_classes', - 'width_coefficient', - 'depth_coefficient', - 'depth_divisor', - 'min_depth', - 'drop_connect_rate', -]) - -BlockArgs = collections.namedtuple('BlockArgs', [ - 'kernel_size', 'num_repeat', 'input_filters', 'output_filters', - 'expand_ratio', 'id_skip', 'stride', 'se_ratio' -]) - -GlobalParams.__new__.__defaults__ = (None, ) * len(GlobalParams._fields) -BlockArgs.__new__.__defaults__ = (None, ) * len(BlockArgs._fields) - - -def efficientnet_params(model_name): - """ Map EfficientNet model name to parameter coefficients. """ - params_dict = { - # Coefficients: width,depth,resolution,dropout - 'efficientnet-b0': (1.0, 1.0, 224, 0.2), - 'efficientnet-b1': (1.0, 1.1, 240, 0.2), - 'efficientnet-b2': (1.1, 1.2, 260, 0.3), - 'efficientnet-b3': (1.2, 1.4, 300, 0.3), - 'efficientnet-b4': (1.4, 1.8, 380, 0.4), - 'efficientnet-b5': (1.6, 2.2, 456, 0.4), - 'efficientnet-b6': (1.8, 2.6, 528, 0.5), - 'efficientnet-b7': (2.0, 3.1, 600, 0.5), - } - return params_dict[model_name] - - -def efficientnet(width_coefficient=None, - depth_coefficient=None, - dropout_rate=0.2, - drop_connect_rate=0.2): - """ Get block arguments according to parameter and coefficients. """ - blocks_args = [ - 'r1_k3_s11_e1_i32_o16_se0.25', - 'r2_k3_s22_e6_i16_o24_se0.25', - 'r2_k5_s22_e6_i24_o40_se0.25', - 'r3_k3_s22_e6_i40_o80_se0.25', - 'r3_k5_s11_e6_i80_o112_se0.25', - 'r4_k5_s22_e6_i112_o192_se0.25', - 'r1_k3_s11_e6_i192_o320_se0.25', - ] - blocks_args = BlockDecoder.decode(blocks_args) - - global_params = GlobalParams( - batch_norm_momentum=0.99, - batch_norm_epsilon=1e-3, - dropout_rate=dropout_rate, - drop_connect_rate=drop_connect_rate, - num_classes=1000, - width_coefficient=width_coefficient, - depth_coefficient=depth_coefficient, - depth_divisor=8, - min_depth=None) - - return blocks_args, global_params - - -def get_model_params(model_name, override_params): - """ Get the block args and global params for a given model """ - if model_name.startswith('efficientnet'): - w, d, _, p = efficientnet_params(model_name) - blocks_args, global_params = efficientnet( - width_coefficient=w, depth_coefficient=d, dropout_rate=p) - else: - raise NotImplementedError( - 'model name is not pre-defined: %s' % model_name) - if override_params: - global_params = global_params._replace(**override_params) - return blocks_args, global_params - - -def round_filters(filters, global_params): - """ Calculate and round number of filters based on depth multiplier. """ - multiplier = global_params.width_coefficient - if not multiplier: - return filters - divisor = global_params.depth_divisor - min_depth = global_params.min_depth - filters *= multiplier - min_depth = min_depth or divisor - new_filters = max(min_depth, - int(filters + divisor / 2) // divisor * divisor) - if new_filters < 0.9 * filters: # prevent rounding by more than 10% - new_filters += divisor - return int(new_filters) - - -def round_repeats(repeats, global_params): - """ Round number of filters based on depth multiplier. """ - multiplier = global_params.depth_coefficient - if not multiplier: - return repeats - return int(math.ceil(multiplier * repeats)) - - -class EfficientNet(): - def __init__(self, - name='b0', - padding_type='SAME', - override_params=None, - is_test=False, - use_se=True): - valid_names = ['b' + str(i) for i in range(8)] - assert name in valid_names, 'efficient name should be in b0~b7' - model_name = 'efficientnet-' + name - self._blocks_args, self._global_params = get_model_params( - model_name, override_params) - self._bn_mom = self._global_params.batch_norm_momentum - self._bn_eps = self._global_params.batch_norm_epsilon - self.padding_type = padding_type - self.use_se = use_se - - def net(self, input, class_dim=1000, is_test=False): - - conv = self.extract_features(input, is_test=is_test) - - out_channels = round_filters(1280, self._global_params) - conv = self.conv_bn_layer( - conv, - num_filters=out_channels, - filter_size=1, - bn_act='swish', - bn_mom=self._bn_mom, - bn_eps=self._bn_eps, - padding_type=self.padding_type, - name='', - conv_name='_conv_head', - bn_name='_bn1') - - pool = fluid.layers.pool2d( - input=conv, pool_type='avg', global_pooling=True, use_cudnn=False) - - if not is_test and self._global_params.dropout_rate: - pool = fluid.layers.dropout( - pool, - self._global_params.dropout_rate, - dropout_implementation='upscale_in_train') - - param_attr, bias_attr = init_fc_layer(class_dim, '_fc') - out = fluid.layers.fc( - pool, - class_dim, - name='_fc', - param_attr=param_attr, - bias_attr=bias_attr) - return out, pool - - def _drop_connect(self, inputs, prob, is_test): - if is_test: - return inputs - keep_prob = 1.0 - prob - random_tensor = keep_prob + fluid.layers.uniform_random_batch_size_like( - inputs, [-1, 1, 1, 1], min=0., max=1.) - binary_tensor = fluid.layers.floor(random_tensor) - output = inputs / keep_prob * binary_tensor - return output - - def _expand_conv_norm(self, inputs, block_args, is_test, name=None): - # Expansion phase - oup = block_args.input_filters * block_args.expand_ratio # number of output channels - - if block_args.expand_ratio != 1: - conv = self.conv_bn_layer( - inputs, - num_filters=oup, - filter_size=1, - bn_act=None, - bn_mom=self._bn_mom, - bn_eps=self._bn_eps, - padding_type=self.padding_type, - name=name, - conv_name=name + '_expand_conv', - bn_name='_bn0') - - return conv - - def _depthwise_conv_norm(self, inputs, block_args, is_test, name=None): - k = block_args.kernel_size - s = block_args.stride - if isinstance(s, list) or isinstance(s, tuple): - s = s[0] - oup = block_args.input_filters * block_args.expand_ratio # number of output channels - - conv = self.conv_bn_layer( - inputs, - num_filters=oup, - filter_size=k, - stride=s, - num_groups=oup, - bn_act=None, - padding_type=self.padding_type, - bn_mom=self._bn_mom, - bn_eps=self._bn_eps, - name=name, - use_cudnn=False, - conv_name=name + '_depthwise_conv', - bn_name='_bn1') - - return conv - - def _project_conv_norm(self, inputs, block_args, is_test, name=None): - final_oup = block_args.output_filters - conv = self.conv_bn_layer( - inputs, - num_filters=final_oup, - filter_size=1, - bn_act=None, - padding_type=self.padding_type, - bn_mom=self._bn_mom, - bn_eps=self._bn_eps, - name=name, - conv_name=name + '_project_conv', - bn_name='_bn2') - return conv - - def conv_bn_layer(self, - input, - filter_size, - num_filters, - stride=1, - num_groups=1, - padding_type="SAME", - conv_act=None, - bn_act='swish', - use_cudnn=True, - use_bn=True, - bn_mom=0.9, - bn_eps=1e-05, - use_bias=False, - name=None, - conv_name=None, - bn_name=None): - conv = conv2d( - input=input, - num_filters=num_filters, - filter_size=filter_size, - stride=stride, - groups=num_groups, - act=conv_act, - padding_type=padding_type, - use_cudnn=use_cudnn, - name=conv_name, - use_bias=use_bias) - - if use_bn == False: - return conv - else: - bn_name = name + bn_name - param_attr, bias_attr = init_batch_norm_layer(bn_name) - return fluid.layers.batch_norm( - input=conv, - act=bn_act, - momentum=bn_mom, - epsilon=bn_eps, - name=bn_name, - moving_mean_name=bn_name + '_mean', - moving_variance_name=bn_name + '_variance', - param_attr=param_attr, - bias_attr=bias_attr) - - def _conv_stem_norm(self, inputs, is_test): - out_channels = round_filters(32, self._global_params) - bn = self.conv_bn_layer( - inputs, - num_filters=out_channels, - filter_size=3, - stride=2, - bn_act=None, - bn_mom=self._bn_mom, - padding_type=self.padding_type, - bn_eps=self._bn_eps, - name='', - conv_name='_conv_stem', - bn_name='_bn0') - - return bn - - def mb_conv_block(self, - inputs, - block_args, - is_test=False, - drop_connect_rate=None, - name=None): - # Expansion and Depthwise Convolution - oup = block_args.input_filters * block_args.expand_ratio # number of output channels - has_se = self.use_se and (block_args.se_ratio is - not None) and (0 < block_args.se_ratio <= 1) - id_skip = block_args.id_skip # skip connection and drop connect - conv = inputs - if block_args.expand_ratio != 1: - conv = fluid.layers.swish( - self._expand_conv_norm(conv, block_args, is_test, name)) - - conv = fluid.layers.swish( - self._depthwise_conv_norm(conv, block_args, is_test, name)) - - # Squeeze and Excitation - if has_se: - num_squeezed_channels = max( - 1, int(block_args.input_filters * block_args.se_ratio)) - conv = self.se_block(conv, num_squeezed_channels, oup, name) - - conv = self._project_conv_norm(conv, block_args, is_test, name) - - # Skip connection and drop connect - input_filters, output_filters = block_args.input_filters, block_args.output_filters - if id_skip and block_args.stride == 1 and input_filters == output_filters: - if drop_connect_rate: - conv = self._drop_connect(conv, drop_connect_rate, is_test) - conv = fluid.layers.elementwise_add(conv, inputs) - - return conv - - def se_block(self, inputs, num_squeezed_channels, oup, name): - x_squeezed = fluid.layers.pool2d( - input=inputs, pool_type='avg', global_pooling=True, use_cudnn=False) - x_squeezed = conv2d( - x_squeezed, - num_filters=num_squeezed_channels, - filter_size=1, - use_bias=True, - padding_type=self.padding_type, - act='swish', - name=name + '_se_reduce') - x_squeezed = conv2d( - x_squeezed, - num_filters=oup, - filter_size=1, - use_bias=True, - padding_type=self.padding_type, - name=name + '_se_expand') - se_out = inputs * fluid.layers.sigmoid(x_squeezed) - return se_out - - def extract_features(self, inputs, is_test): - """ Returns output of the final convolution layer """ - - conv = fluid.layers.swish(self._conv_stem_norm(inputs, is_test=is_test)) - - block_args_copy = copy.deepcopy(self._blocks_args) - idx = 0 - block_size = 0 - for block_arg in block_args_copy: - block_arg = block_arg._replace( - input_filters=round_filters(block_arg.input_filters, - self._global_params), - output_filters=round_filters(block_arg.output_filters, - self._global_params), - num_repeat=round_repeats(block_arg.num_repeat, - self._global_params)) - block_size += 1 - for _ in range(block_arg.num_repeat - 1): - block_size += 1 - - for block_args in self._blocks_args: - - # Update block input and output filters based on depth multiplier. - block_args = block_args._replace( - input_filters=round_filters(block_args.input_filters, - self._global_params), - output_filters=round_filters(block_args.output_filters, - self._global_params), - num_repeat=round_repeats(block_args.num_repeat, - self._global_params)) - - # The first block needs to take care of stride and filter size increase. - drop_connect_rate = self._global_params.drop_connect_rate - if drop_connect_rate: - drop_connect_rate *= float(idx) / block_size - conv = self.mb_conv_block(conv, block_args, is_test, - drop_connect_rate, - '_blocks.' + str(idx) + '.') - - idx += 1 - if block_args.num_repeat > 1: - block_args = block_args._replace( - input_filters=block_args.output_filters, stride=1) - for _ in range(block_args.num_repeat - 1): - drop_connect_rate = self._global_params.drop_connect_rate - if drop_connect_rate: - drop_connect_rate *= float(idx) / block_size - conv = self.mb_conv_block(conv, block_args, is_test, - drop_connect_rate, - '_blocks.' + str(idx) + '.') - idx += 1 - - return conv - - def shortcut(self, input, data_residual): - return fluid.layers.elementwise_add(input, data_residual) - - -class BlockDecoder(object): - """ Block Decoder for readability, straight from the official TensorFlow repository """ - - @staticmethod - def _decode_block_string(block_string): - """ Gets a block through a string notation of arguments. """ - assert isinstance(block_string, str) - - ops = block_string.split('_') - options = {} - for op in ops: - splits = re.split(r'(\d.*)', op) - if len(splits) >= 2: - key, value = splits[:2] - options[key] = value - - # Check stride - assert (('s' in options and len(options['s']) == 1) or - (len(options['s']) == 2 and options['s'][0] == options['s'][1])) - - return BlockArgs( - kernel_size=int(options['k']), - num_repeat=int(options['r']), - input_filters=int(options['i']), - output_filters=int(options['o']), - expand_ratio=int(options['e']), - id_skip=('noskip' not in block_string), - se_ratio=float(options['se']) if 'se' in options else None, - stride=[int(options['s'][0])]) - - @staticmethod - def _encode_block_string(block): - """Encodes a block to a string.""" - args = [ - 'r%d' % block.num_repeat, - 'k%d' % block.kernel_size, - 's%d%d' % (block.strides[0], block.strides[1]), - 'e%s' % block.expand_ratio, - 'i%d' % block.input_filters, - 'o%d' % block.output_filters - ] - if 0 < block.se_ratio <= 1: - args.append('se%s' % block.se_ratio) - if block.id_skip is False: - args.append('noskip') - return '_'.join(args) - - @staticmethod - def decode(string_list): - """ - Decodes a list of string notations to specify blocks inside the network. - :param string_list: a list of strings, each string is a notation of block - :return: a list of BlockArgs namedtuples of block args - """ - assert isinstance(string_list, list) - blocks_args = [] - for block_string in string_list: - blocks_args.append(BlockDecoder._decode_block_string(block_string)) - return blocks_args - - @staticmethod - def encode(blocks_args): - """ - Encodes a list of BlockArgs to a list of strings. - :param blocks_args: a list of BlockArgs namedtuples of block args - :return: a list of strings, each string is a notation of block - """ - block_strings = [] - for block in blocks_args: - block_strings.append(BlockDecoder._encode_block_string(block)) - return block_strings - - -def EfficientNetB0_small(is_test=False, - padding_type='SAME', - override_params=None, - use_se=False): - model = EfficientNet( - name='b0', - is_test=is_test, - padding_type=padding_type, - override_params=override_params, - use_se=use_se) - return model - - -def EfficientNetB0(is_test=False, - padding_type='SAME', - override_params=None, - use_se=True): - model = EfficientNet( - name='b0', - is_test=is_test, - padding_type=padding_type, - override_params=override_params, - use_se=use_se) - return model - - -def EfficientNetB1(is_test=False, - padding_type='SAME', - override_params=None, - use_se=True): - model = EfficientNet( - name='b1', - is_test=is_test, - padding_type=padding_type, - override_params=override_params, - use_se=use_se) - return model - - -def EfficientNetB2(is_test=False, - padding_type='SAME', - override_params=None, - use_se=True): - model = EfficientNet( - name='b2', - is_test=is_test, - padding_type=padding_type, - override_params=override_params, - use_se=use_se) - return model - - -def EfficientNetB3(is_test=False, - padding_type='SAME', - override_params=None, - use_se=True): - model = EfficientNet( - name='b3', - is_test=is_test, - padding_type=padding_type, - override_params=override_params, - use_se=use_se) - return model - - -def EfficientNetB4(is_test=False, - padding_type='SAME', - override_params=None, - use_se=True): - model = EfficientNet( - name='b4', - is_test=is_test, - padding_type=padding_type, - override_params=override_params, - use_se=use_se) - return model - - -def EfficientNetB5(is_test=False, - padding_type='SAME', - override_params=None, - use_se=True): - model = EfficientNet( - name='b5', - is_test=is_test, - padding_type=padding_type, - override_params=override_params, - use_se=use_se) - return model - - -def EfficientNetB6(is_test=False, - padding_type='SAME', - override_params=None, - use_se=True): - model = EfficientNet( - name='b6', - is_test=is_test, - padding_type=padding_type, - override_params=override_params, - use_se=use_se) - return model - - -def EfficientNetB7(is_test=False, - padding_type='SAME', - override_params=None, - use_se=True): - model = EfficientNet( - name='b7', - is_test=is_test, - padding_type=padding_type, - override_params=override_params, - use_se=use_se) - return model diff --git a/hub_module/modules/image/classification/efficientnetb1_imagenet/label_list.txt b/hub_module/modules/image/classification/efficientnetb1_imagenet/label_list.txt deleted file mode 100644 index a509c007..00000000 --- a/hub_module/modules/image/classification/efficientnetb1_imagenet/label_list.txt +++ /dev/null @@ -1,1000 +0,0 @@ -tench, Tinca tinca -goldfish, Carassius auratus -great white shark, white shark, man-eater, man-eating shark, Carcharodon carcharias -tiger shark, Galeocerdo cuvieri -hammerhead, hammerhead shark -electric ray, crampfish, numbfish, torpedo -stingray -cock -hen -ostrich, Struthio camelus -brambling, Fringilla montifringilla -goldfinch, Carduelis carduelis -house finch, linnet, Carpodacus mexicanus -junco, snowbird -indigo bunting, indigo finch, indigo bird, Passerina cyanea -robin, American robin, Turdus migratorius -bulbul -jay -magpie -chickadee -water ouzel, dipper -kite -bald eagle, American eagle, Haliaeetus leucocephalus -vulture -great grey owl, great gray owl, Strix nebulosa -European fire salamander, Salamandra salamandra -common newt, Triturus vulgaris -eft -spotted salamander, Ambystoma maculatum -axolotl, mud puppy, Ambystoma mexicanum -bullfrog, Rana catesbeiana -tree frog, tree-frog -tailed frog, bell toad, ribbed toad, tailed toad, Ascaphus trui -loggerhead, loggerhead turtle, Caretta caretta -leatherback turtle, leatherback, leathery turtle, Dermochelys coriacea -mud turtle -terrapin -box turtle, box tortoise -banded gecko -common iguana, iguana, Iguana iguana -American chameleon, anole, Anolis carolinensis -whiptail, whiptail lizard -agama -frilled lizard, Chlamydosaurus kingi -alligator lizard -Gila monster, Heloderma suspectum -green lizard, Lacerta viridis -African chameleon, Chamaeleo chamaeleon -Komodo dragon, Komodo lizard, dragon lizard, giant lizard, Varanus komodoensis -African crocodile, Nile crocodile, Crocodylus niloticus -American alligator, Alligator mississipiensis -triceratops -thunder snake, worm snake, Carphophis amoenus -ringneck snake, ring-necked snake, ring snake -hognose snake, puff adder, sand viper -green snake, grass snake -king snake, kingsnake -garter snake, grass snake -water snake -vine snake -night snake, Hypsiglena torquata -boa constrictor, Constrictor constrictor -rock python, rock snake, Python sebae -Indian cobra, Naja naja -green mamba -sea snake -horned viper, cerastes, sand viper, horned asp, Cerastes cornutus -diamondback, diamondback rattlesnake, Crotalus adamanteus -sidewinder, horned rattlesnake, Crotalus cerastes -trilobite -harvestman, daddy longlegs, Phalangium opilio -scorpion -black and gold garden spider, Argiope aurantia -barn spider, Araneus cavaticus -garden spider, Aranea diademata -black widow, Latrodectus mactans -tarantula -wolf spider, hunting spider -tick -centipede -black grouse -ptarmigan -ruffed grouse, partridge, Bonasa umbellus -prairie chicken, prairie grouse, prairie fowl -peacock -quail -partridge -African grey, African gray, Psittacus erithacus -macaw -sulphur-crested cockatoo, Kakatoe galerita, Cacatua galerita -lorikeet -coucal -bee eater -hornbill -hummingbird -jacamar -toucan -drake -red-breasted merganser, Mergus serrator -goose -black swan, Cygnus atratus -tusker -echidna, spiny anteater, anteater -platypus, duckbill, duckbilled platypus, duck-billed platypus, Ornithorhynchus anatinus -wallaby, brush kangaroo -koala, koala bear, kangaroo bear, native bear, Phascolarctos cinereus -wombat -jellyfish -sea anemone, anemone -brain coral -flatworm, platyhelminth -nematode, nematode worm, roundworm -conch -snail -slug -sea slug, nudibranch -chiton, coat-of-mail shell, sea cradle, polyplacophore -chambered nautilus, pearly nautilus, nautilus -Dungeness crab, Cancer magister -rock crab, Cancer irroratus -fiddler crab -king crab, Alaska crab, Alaskan king crab, Alaska king crab, Paralithodes camtschatica -American lobster, Northern lobster, Maine lobster, Homarus americanus -spiny lobster, langouste, rock lobster, crawfish, crayfish, sea crawfish -crayfish, crawfish, crawdad, crawdaddy -hermit crab -isopod -white stork, Ciconia ciconia -black stork, Ciconia nigra -spoonbill -flamingo -little blue heron, Egretta caerulea -American egret, great white heron, Egretta albus -bittern -crane -limpkin, Aramus pictus -European gallinule, Porphyrio porphyrio -American coot, marsh hen, mud hen, water hen, Fulica americana -bustard -ruddy turnstone, Arenaria interpres -red-backed sandpiper, dunlin, Erolia alpina -redshank, Tringa totanus -dowitcher -oystercatcher, oyster catcher -pelican -king penguin, Aptenodytes patagonica -albatross, mollymawk -grey whale, gray whale, devilfish, Eschrichtius gibbosus, Eschrichtius robustus -killer whale, killer, orca, grampus, sea wolf, Orcinus orca -dugong, Dugong dugon -sea lion -Chihuahua -Japanese spaniel -Maltese dog, Maltese terrier, Maltese -Pekinese, Pekingese, Peke -Shih-Tzu -Blenheim spaniel -papillon -toy terrier -Rhodesian ridgeback -Afghan hound, Afghan -basset, basset hound -beagle -bloodhound, sleuthhound -bluetick -black-and-tan coonhound -Walker hound, Walker foxhound -English foxhound -redbone -borzoi, Russian wolfhound -Irish wolfhound -Italian greyhound -whippet -Ibizan hound, Ibizan Podenco -Norwegian elkhound, elkhound -otterhound, otter hound -Saluki, gazelle hound -Scottish deerhound, deerhound -Weimaraner -Staffordshire bullterrier, Staffordshire bull terrier -American Staffordshire terrier, Staffordshire terrier, American pit bull terrier, pit bull terrier -Bedlington terrier -Border terrier -Kerry blue terrier -Irish terrier -Norfolk terrier -Norwich terrier -Yorkshire terrier -wire-haired fox terrier -Lakeland terrier -Sealyham terrier, Sealyham -Airedale, Airedale terrier -cairn, cairn terrier -Australian terrier -Dandie Dinmont, Dandie Dinmont terrier -Boston bull, Boston terrier -miniature schnauzer -giant schnauzer -standard schnauzer -Scotch terrier, Scottish terrier, Scottie -Tibetan terrier, chrysanthemum dog -silky terrier, Sydney silky -soft-coated wheaten terrier -West Highland white terrier -Lhasa, Lhasa apso -flat-coated retriever -curly-coated retriever -golden retriever -Labrador retriever -Chesapeake Bay retriever -German short-haired pointer -vizsla, Hungarian pointer -English setter -Irish setter, red setter -Gordon setter -Brittany spaniel -clumber, clumber spaniel -English springer, English springer spaniel -Welsh springer spaniel -cocker spaniel, English cocker spaniel, cocker -Sussex spaniel -Irish water spaniel -kuvasz -schipperke -groenendael -malinois -briard -kelpie -komondor -Old English sheepdog, bobtail -Shetland sheepdog, Shetland sheep dog, Shetland -collie -Border collie -Bouvier des Flandres, Bouviers des Flandres -Rottweiler -German shepherd, German shepherd dog, German police dog, alsatian -Doberman, Doberman pinscher -miniature pinscher -Greater Swiss Mountain dog -Bernese mountain dog -Appenzeller -EntleBucher -boxer -bull mastiff -Tibetan mastiff -French bulldog -Great Dane -Saint Bernard, St Bernard -Eskimo dog, husky -malamute, malemute, Alaskan malamute -Siberian husky -dalmatian, coach dog, carriage dog -affenpinscher, monkey pinscher, monkey dog -basenji -pug, pug-dog -Leonberg -Newfoundland, Newfoundland dog -Great Pyrenees -Samoyed, Samoyede -Pomeranian -chow, chow chow -keeshond -Brabancon griffon -Pembroke, Pembroke Welsh corgi -Cardigan, Cardigan Welsh corgi -toy poodle -miniature poodle -standard poodle -Mexican hairless -timber wolf, grey wolf, gray wolf, Canis lupus -white wolf, Arctic wolf, Canis lupus tundrarum -red wolf, maned wolf, Canis rufus, Canis niger -coyote, prairie wolf, brush wolf, Canis latrans -dingo, warrigal, warragal, Canis dingo -dhole, Cuon alpinus -African hunting dog, hyena dog, Cape hunting dog, Lycaon pictus -hyena, hyaena -red fox, Vulpes vulpes -kit fox, Vulpes macrotis -Arctic fox, white fox, Alopex lagopus -grey fox, gray fox, Urocyon cinereoargenteus -tabby, tabby cat -tiger cat -Persian cat -Siamese cat, Siamese -Egyptian cat -cougar, puma, catamount, mountain lion, painter, panther, Felis concolor -lynx, catamount -leopard, Panthera pardus -snow leopard, ounce, Panthera uncia -jaguar, panther, Panthera onca, Felis onca -lion, king of beasts, Panthera leo -tiger, Panthera tigris -cheetah, chetah, Acinonyx jubatus -brown bear, bruin, Ursus arctos -American black bear, black bear, Ursus americanus, Euarctos americanus -ice bear, polar bear, Ursus Maritimus, Thalarctos maritimus -sloth bear, Melursus ursinus, Ursus ursinus -mongoose -meerkat, mierkat -tiger beetle -ladybug, ladybeetle, lady beetle, ladybird, ladybird beetle -ground beetle, carabid beetle -long-horned beetle, longicorn, longicorn beetle -leaf beetle, chrysomelid -dung beetle -rhinoceros beetle -weevil -fly -bee -ant, emmet, pismire -grasshopper, hopper -cricket -walking stick, walkingstick, stick insect -cockroach, roach -mantis, mantid -cicada, cicala -leafhopper -lacewing, lacewing fly -dragonfly, darning needle, devil's darning needle, sewing needle, snake feeder, snake doctor, mosquito hawk, skeeter hawk -damselfly -admiral -ringlet, ringlet butterfly -monarch, monarch butterfly, milkweed butterfly, Danaus plexippus -cabbage butterfly -sulphur butterfly, sulfur butterfly -lycaenid, lycaenid butterfly -starfish, sea star -sea urchin -sea cucumber, holothurian -wood rabbit, cottontail, cottontail rabbit -hare -Angora, Angora rabbit -hamster -porcupine, hedgehog -fox squirrel, eastern fox squirrel, Sciurus niger -marmot -beaver -guinea pig, Cavia cobaya -sorrel -zebra -hog, pig, grunter, squealer, Sus scrofa -wild boar, boar, Sus scrofa -warthog -hippopotamus, hippo, river horse, Hippopotamus amphibius -ox -water buffalo, water ox, Asiatic buffalo, Bubalus bubalis -bison -ram, tup -bighorn, bighorn sheep, cimarron, Rocky Mountain bighorn, Rocky Mountain sheep, Ovis canadensis -ibex, Capra ibex -hartebeest -impala, Aepyceros melampus -gazelle -Arabian camel, dromedary, Camelus dromedarius -llama -weasel -mink -polecat, fitch, foulmart, foumart, Mustela putorius -black-footed ferret, ferret, Mustela nigripes -otter -skunk, polecat, wood pussy -badger -armadillo -three-toed sloth, ai, Bradypus tridactylus -orangutan, orang, orangutang, Pongo pygmaeus -gorilla, Gorilla gorilla -chimpanzee, chimp, Pan troglodytes -gibbon, Hylobates lar -siamang, Hylobates syndactylus, Symphalangus syndactylus -guenon, guenon monkey -patas, hussar monkey, Erythrocebus patas -baboon -macaque -langur -colobus, colobus monkey -proboscis monkey, Nasalis larvatus -marmoset -capuchin, ringtail, Cebus capucinus -howler monkey, howler -titi, titi monkey -spider monkey, Ateles geoffroyi -squirrel monkey, Saimiri sciureus -Madagascar cat, ring-tailed lemur, Lemur catta -indri, indris, Indri indri, Indri brevicaudatus -Indian elephant, Elephas maximus -African elephant, Loxodonta africana -lesser panda, red panda, panda, bear cat, cat bear, Ailurus fulgens -giant panda, panda, panda bear, coon bear, Ailuropoda melanoleuca -barracouta, snoek -eel -coho, cohoe, coho salmon, blue jack, silver salmon, Oncorhynchus kisutch -rock beauty, Holocanthus tricolor -anemone fish -sturgeon -gar, garfish, garpike, billfish, Lepisosteus osseus -lionfish -puffer, pufferfish, blowfish, globefish -abacus -abaya -academic gown, academic robe, judge's robe -accordion, piano accordion, squeeze box -acoustic guitar -aircraft carrier, carrier, flattop, attack aircraft carrier -airliner -airship, dirigible -altar -ambulance -amphibian, amphibious vehicle -analog clock -apiary, bee house -apron -ashcan, trash can, garbage can, wastebin, ash bin, ash-bin, ashbin, dustbin, trash barrel, trash bin -assault rifle, assault gun -backpack, back pack, knapsack, packsack, rucksack, haversack -bakery, bakeshop, bakehouse -balance beam, beam -balloon -ballpoint, ballpoint pen, ballpen, Biro -Band Aid -banjo -bannister, banister, balustrade, balusters, handrail -barbell -barber chair -barbershop -barn -barometer -barrel, cask -barrow, garden cart, lawn cart, wheelbarrow -baseball -basketball -bassinet -bassoon -bathing cap, swimming cap -bath towel -bathtub, bathing tub, bath, tub -beach wagon, station wagon, wagon, estate car, beach waggon, station waggon, waggon -beacon, lighthouse, beacon light, pharos -beaker -bearskin, busby, shako -beer bottle -beer glass -bell cote, bell cot -bib -bicycle-built-for-two, tandem bicycle, tandem -bikini, two-piece -binder, ring-binder -binoculars, field glasses, opera glasses -birdhouse -boathouse -bobsled, bobsleigh, bob -bolo tie, bolo, bola tie, bola -bonnet, poke bonnet -bookcase -bookshop, bookstore, bookstall -bottlecap -bow -bow tie, bow-tie, bowtie -brass, memorial tablet, plaque -brassiere, bra, bandeau -breakwater, groin, groyne, mole, bulwark, seawall, jetty -breastplate, aegis, egis -broom -bucket, pail -buckle -bulletproof vest -bullet train, bullet -butcher shop, meat market -cab, hack, taxi, taxicab -caldron, cauldron -candle, taper, wax light -cannon -canoe -can opener, tin opener -cardigan -car mirror -carousel, carrousel, merry-go-round, roundabout, whirligig -carpenter's kit, tool kit -carton -car wheel -cash machine, cash dispenser, automated teller machine, automatic teller machine, automated teller, automatic teller, ATM -cassette -cassette player -castle -catamaran -CD player -cello, violoncello -cellular telephone, cellular phone, cellphone, cell, mobile phone -chain -chainlink fence -chain mail, ring mail, mail, chain armor, chain armour, ring armor, ring armour -chain saw, chainsaw -chest -chiffonier, commode -chime, bell, gong -china cabinet, china closet -Christmas stocking -church, church building -cinema, movie theater, movie theatre, movie house, picture palace -cleaver, meat cleaver, chopper -cliff dwelling -cloak -clog, geta, patten, sabot -cocktail shaker -coffee mug -coffeepot -coil, spiral, volute, whorl, helix -combination lock -computer keyboard, keypad -confectionery, confectionary, candy store -container ship, containership, container vessel -convertible -corkscrew, bottle screw -cornet, horn, trumpet, trump -cowboy boot -cowboy hat, ten-gallon hat -cradle -crane -crash helmet -crate -crib, cot -Crock Pot -croquet ball -crutch -cuirass -dam, dike, dyke -desk -desktop computer -dial telephone, dial phone -diaper, nappy, napkin -digital clock -digital watch -dining table, board -dishrag, dishcloth -dishwasher, dish washer, dishwashing machine -disk brake, disc brake -dock, dockage, docking facility -dogsled, dog sled, dog sleigh -dome -doormat, welcome mat -drilling platform, offshore rig -drum, membranophone, tympan -drumstick -dumbbell -Dutch oven -electric fan, blower -electric guitar -electric locomotive -entertainment center -envelope -espresso maker -face powder -feather boa, boa -file, file cabinet, filing cabinet -fireboat -fire engine, fire truck -fire screen, fireguard -flagpole, flagstaff -flute, transverse flute -folding chair -football helmet -forklift -fountain -fountain pen -four-poster -freight car -French horn, horn -frying pan, frypan, skillet -fur coat -garbage truck, dustcart -gasmask, respirator, gas helmet -gas pump, gasoline pump, petrol pump, island dispenser -goblet -go-kart -golf ball -golfcart, golf cart -gondola -gong, tam-tam -gown -grand piano, grand -greenhouse, nursery, glasshouse -grille, radiator grille -grocery store, grocery, food market, market -guillotine -hair slide -hair spray -half track -hammer -hamper -hand blower, blow dryer, blow drier, hair dryer, hair drier -hand-held computer, hand-held microcomputer -handkerchief, hankie, hanky, hankey -hard disc, hard disk, fixed disk -harmonica, mouth organ, harp, mouth harp -harp -harvester, reaper -hatchet -holster -home theater, home theatre -honeycomb -hook, claw -hoopskirt, crinoline -horizontal bar, high bar -horse cart, horse-cart -hourglass -iPod -iron, smoothing iron -jack-o'-lantern -jean, blue jean, denim -jeep, landrover -jersey, T-shirt, tee shirt -jigsaw puzzle -jinrikisha, ricksha, rickshaw -joystick -kimono -knee pad -knot -lab coat, laboratory coat -ladle -lampshade, lamp shade -laptop, laptop computer -lawn mower, mower -lens cap, lens cover -letter opener, paper knife, paperknife -library -lifeboat -lighter, light, igniter, ignitor -limousine, limo -liner, ocean liner -lipstick, lip rouge -Loafer -lotion -loudspeaker, speaker, speaker unit, loudspeaker system, speaker system -loupe, jeweler's loupe -lumbermill, sawmill -magnetic compass -mailbag, postbag -mailbox, letter box -maillot -maillot, tank suit -manhole cover -maraca -marimba, xylophone -mask -matchstick -maypole -maze, labyrinth -measuring cup -medicine chest, medicine cabinet -megalith, megalithic structure -microphone, mike -microwave, microwave oven -military uniform -milk can -minibus -miniskirt, mini -minivan -missile -mitten -mixing bowl -mobile home, manufactured home -Model T -modem -monastery -monitor -moped -mortar -mortarboard -mosque -mosquito net -motor scooter, scooter -mountain bike, all-terrain bike, off-roader -mountain tent -mouse, computer mouse -mousetrap -moving van -muzzle -nail -neck brace -necklace -nipple -notebook, notebook computer -obelisk -oboe, hautboy, hautbois -ocarina, sweet potato -odometer, hodometer, mileometer, milometer -oil filter -organ, pipe organ -oscilloscope, scope, cathode-ray oscilloscope, CRO -overskirt -oxcart -oxygen mask -packet -paddle, boat paddle -paddlewheel, paddle wheel -padlock -paintbrush -pajama, pyjama, pj's, jammies -palace -panpipe, pandean pipe, syrinx -paper towel -parachute, chute -parallel bars, bars -park bench -parking meter -passenger car, coach, carriage -patio, terrace -pay-phone, pay-station -pedestal, plinth, footstall -pencil box, pencil case -pencil sharpener -perfume, essence -Petri dish -photocopier -pick, plectrum, plectron -pickelhaube -picket fence, paling -pickup, pickup truck -pier -piggy bank, penny bank -pill bottle -pillow -ping-pong ball -pinwheel -pirate, pirate ship -pitcher, ewer -plane, carpenter's plane, woodworking plane -planetarium -plastic bag -plate rack -plow, plough -plunger, plumber's helper -Polaroid camera, Polaroid Land camera -pole -police van, police wagon, paddy wagon, patrol wagon, wagon, black Maria -poncho -pool table, billiard table, snooker table -pop bottle, soda bottle -pot, flowerpot -potter's wheel -power drill -prayer rug, prayer mat -printer -prison, prison house -projectile, missile -projector -puck, hockey puck -punching bag, punch bag, punching ball, punchball -purse -quill, quill pen -quilt, comforter, comfort, puff -racer, race car, racing car -racket, racquet -radiator -radio, wireless -radio telescope, radio reflector -rain barrel -recreational vehicle, RV, R.V. -reel -reflex camera -refrigerator, icebox -remote control, remote -restaurant, eating house, eating place, eatery -revolver, six-gun, six-shooter -rifle -rocking chair, rocker -rotisserie -rubber eraser, rubber, pencil eraser -rugby ball -rule, ruler -running shoe -safe -safety pin -saltshaker, salt shaker -sandal -sarong -sax, saxophone -scabbard -scale, weighing machine -school bus -schooner -scoreboard -screen, CRT screen -screw -screwdriver -seat belt, seatbelt -sewing machine -shield, buckler -shoe shop, shoe-shop, shoe store -shoji -shopping basket -shopping cart -shovel -shower cap -shower curtain -ski -ski mask -sleeping bag -slide rule, slipstick -sliding door -slot, one-armed bandit -snorkel -snowmobile -snowplow, snowplough -soap dispenser -soccer ball -sock -solar dish, solar collector, solar furnace -sombrero -soup bowl -space bar -space heater -space shuttle -spatula -speedboat -spider web, spider's web -spindle -sports car, sport car -spotlight, spot -stage -steam locomotive -steel arch bridge -steel drum -stethoscope -stole -stone wall -stopwatch, stop watch -stove -strainer -streetcar, tram, tramcar, trolley, trolley car -stretcher -studio couch, day bed -stupa, tope -submarine, pigboat, sub, U-boat -suit, suit of clothes -sundial -sunglass -sunglasses, dark glasses, shades -sunscreen, sunblock, sun blocker -suspension bridge -swab, swob, mop -sweatshirt -swimming trunks, bathing trunks -swing -switch, electric switch, electrical switch -syringe -table lamp -tank, army tank, armored combat vehicle, armoured combat vehicle -tape player -teapot -teddy, teddy bear -television, television system -tennis ball -thatch, thatched roof -theater curtain, theatre curtain -thimble -thresher, thrasher, threshing machine -throne -tile roof -toaster -tobacco shop, tobacconist shop, tobacconist -toilet seat -torch -totem pole -tow truck, tow car, wrecker -toyshop -tractor -trailer truck, tractor trailer, trucking rig, rig, articulated lorry, semi -tray -trench coat -tricycle, trike, velocipede -trimaran -tripod -triumphal arch -trolleybus, trolley coach, trackless trolley -trombone -tub, vat -turnstile -typewriter keyboard -umbrella -unicycle, monocycle -upright, upright piano -vacuum, vacuum cleaner -vase -vault -velvet -vending machine -vestment -viaduct -violin, fiddle -volleyball -waffle iron -wall clock -wallet, billfold, notecase, pocketbook -wardrobe, closet, press -warplane, military plane -washbasin, handbasin, washbowl, lavabo, wash-hand basin -washer, automatic washer, washing machine -water bottle -water jug -water tower -whiskey jug -whistle -wig -window screen -window shade -Windsor tie -wine bottle -wing -wok -wooden spoon -wool, woolen, woollen -worm fence, snake fence, snake-rail fence, Virginia fence -wreck -yawl -yurt -web site, website, internet site, site -comic book -crossword puzzle, crossword -street sign -traffic light, traffic signal, stoplight -book jacket, dust cover, dust jacket, dust wrapper -menu -plate -guacamole -consomme -hot pot, hotpot -trifle -ice cream, icecream -ice lolly, lolly, lollipop, popsicle -French loaf -bagel, beigel -pretzel -cheeseburger -hotdog, hot dog, red hot -mashed potato -head cabbage -broccoli -cauliflower -zucchini, courgette -spaghetti squash -acorn squash -butternut squash -cucumber, cuke -artichoke, globe artichoke -bell pepper -cardoon -mushroom -Granny Smith -strawberry -orange -lemon -fig -pineapple, ananas -banana -jackfruit, jak, jack -custard apple -pomegranate -hay -carbonara -chocolate sauce, chocolate syrup -dough -meat loaf, meatloaf -pizza, pizza pie -potpie -burrito -red wine -espresso -cup -eggnog -alp -bubble -cliff, drop, drop-off -coral reef -geyser -lakeside, lakeshore -promontory, headland, head, foreland -sandbar, sand bar -seashore, coast, seacoast, sea-coast -valley, vale -volcano -ballplayer, baseball player -groom, bridegroom -scuba diver -rapeseed -daisy -yellow lady's slipper, yellow lady-slipper, Cypripedium calceolus, Cypripedium parviflorum -corn -acorn -hip, rose hip, rosehip -buckeye, horse chestnut, conker -coral fungus -agaric -gyromitra -stinkhorn, carrion fungus -earthstar -hen-of-the-woods, hen of the woods, Polyporus frondosus, Grifola frondosa -bolete -ear, spike, capitulum -toilet tissue, toilet paper, bathroom tissue diff --git a/hub_module/modules/image/classification/efficientnetb1_imagenet/layers.py b/hub_module/modules/image/classification/efficientnetb1_imagenet/layers.py deleted file mode 100644 index 00f792b5..00000000 --- a/hub_module/modules/image/classification/efficientnetb1_imagenet/layers.py +++ /dev/null @@ -1,248 +0,0 @@ -# -*- coding:utf-8 -*- -# copyright (c) 2020 PaddlePaddle Authors. All Rights Reserve. -# -# Licensed under the Apache License, Version 2.0 (the "License"); -# you may not use this file except in compliance with the License. -# You may obtain a copy of the License at -# -# http://www.apache.org/licenses/LICENSE-2.0 -# -# Unless required by applicable law or agreed to in writing, software -# distributed under the License is distributed on an "AS IS" BASIS, -# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. -# See the License for the specific language governing permissions and -# limitations under the License. - -from __future__ import absolute_import -from __future__ import division -from __future__ import print_function - -import math -import warnings - -import paddle.fluid as fluid - - -def initial_type(name, - input, - op_type, - fan_out, - init="google", - use_bias=False, - filter_size=0, - stddev=0.02): - if init == "kaiming": - if op_type == 'conv': - fan_in = input.shape[1] * filter_size * filter_size - elif op_type == 'deconv': - fan_in = fan_out * filter_size * filter_size - else: - if len(input.shape) > 2: - fan_in = input.shape[1] * input.shape[2] * input.shape[3] - else: - fan_in = input.shape[1] - bound = 1 / math.sqrt(fan_in) - param_attr = fluid.ParamAttr( - name=name + "_weights", - initializer=fluid.initializer.Uniform(low=-bound, high=bound)) - if use_bias == True: - bias_attr = fluid.ParamAttr( - name=name + '_offset', - initializer=fluid.initializer.Uniform(low=-bound, high=bound)) - else: - bias_attr = False - elif init == 'google': - n = filter_size * filter_size * fan_out - param_attr = fluid.ParamAttr( - name=name + "_weights", - initializer=fluid.initializer.NormalInitializer( - loc=0.0, scale=math.sqrt(2.0 / n))) - if use_bias == True: - bias_attr = fluid.ParamAttr( - name=name + "_offset", - initializer=fluid.initializer.Constant(0.0)) - else: - bias_attr = False - - else: - param_attr = fluid.ParamAttr( - name=name + "_weights", - initializer=fluid.initializer.NormalInitializer( - loc=0.0, scale=stddev)) - if use_bias == True: - bias_attr = fluid.ParamAttr( - name=name + "_offset", - initializer=fluid.initializer.Constant(0.0)) - else: - bias_attr = False - return param_attr, bias_attr - - -def cal_padding(img_size, stride, filter_size, dilation=1): - """Calculate padding size.""" - if img_size % stride == 0: - out_size = max(filter_size - stride, 0) - else: - out_size = max(filter_size - (img_size % stride), 0) - return out_size // 2, out_size - out_size // 2 - - -def init_batch_norm_layer(name="batch_norm"): - param_attr = fluid.ParamAttr( - name=name + '_scale', initializer=fluid.initializer.Constant(1.0)) - bias_attr = fluid.ParamAttr( - name=name + '_offset', - initializer=fluid.initializer.Constant(value=0.0)) - return param_attr, bias_attr - - -def init_fc_layer(fout, name='fc'): - n = fout # fan-out - init_range = 1.0 / math.sqrt(n) - - param_attr = fluid.ParamAttr( - name=name + '_weights', - initializer=fluid.initializer.UniformInitializer( - low=-init_range, high=init_range)) - bias_attr = fluid.ParamAttr( - name=name + '_offset', - initializer=fluid.initializer.Constant(value=0.0)) - return param_attr, bias_attr - - -def norm_layer(input, norm_type='batch_norm', name=None): - if norm_type == 'batch_norm': - param_attr = fluid.ParamAttr( - name=name + '_weights', initializer=fluid.initializer.Constant(1.0)) - bias_attr = fluid.ParamAttr( - name=name + '_offset', - initializer=fluid.initializer.Constant(value=0.0)) - return fluid.layers.batch_norm( - input, - param_attr=param_attr, - bias_attr=bias_attr, - moving_mean_name=name + '_mean', - moving_variance_name=name + '_variance') - - elif norm_type == 'instance_norm': - helper = fluid.layer_helper.LayerHelper("instance_norm", **locals()) - dtype = helper.input_dtype() - epsilon = 1e-5 - mean = fluid.layers.reduce_mean(input, dim=[2, 3], keep_dim=True) - var = fluid.layers.reduce_mean( - fluid.layers.square(input - mean), dim=[2, 3], keep_dim=True) - if name is not None: - scale_name = name + "_scale" - offset_name = name + "_offset" - scale_param = fluid.ParamAttr( - name=scale_name, - initializer=fluid.initializer.Constant(1.0), - trainable=True) - offset_param = fluid.ParamAttr( - name=offset_name, - initializer=fluid.initializer.Constant(0.0), - trainable=True) - scale = helper.create_parameter( - attr=scale_param, shape=input.shape[1:2], dtype=dtype) - offset = helper.create_parameter( - attr=offset_param, shape=input.shape[1:2], dtype=dtype) - - tmp = fluid.layers.elementwise_mul(x=(input - mean), y=scale, axis=1) - tmp = tmp / fluid.layers.sqrt(var + epsilon) - tmp = fluid.layers.elementwise_add(tmp, offset, axis=1) - return tmp - else: - raise NotImplementedError("norm tyoe: [%s] is not support" % norm_type) - - -def conv2d(input, - num_filters=64, - filter_size=7, - stride=1, - stddev=0.02, - padding=0, - groups=None, - name="conv2d", - norm=None, - act=None, - relufactor=0.0, - use_bias=False, - padding_type=None, - initial="normal", - use_cudnn=True): - - if padding != 0 and padding_type != None: - warnings.warn( - 'padding value and padding type are set in the same time, and the final padding width and padding height are computed by padding_type' - ) - - param_attr, bias_attr = initial_type( - name=name, - input=input, - op_type='conv', - fan_out=num_filters, - init=initial, - use_bias=use_bias, - filter_size=filter_size, - stddev=stddev) - - def get_padding(filter_size, stride=1, dilation=1): - padding = ((stride - 1) + dilation * (filter_size - 1)) // 2 - return padding - - need_crop = False - if padding_type == "SAME": - top_padding, bottom_padding = cal_padding(input.shape[2], stride, - filter_size) - left_padding, right_padding = cal_padding(input.shape[2], stride, - filter_size) - height_padding = bottom_padding - width_padding = right_padding - if top_padding != bottom_padding or left_padding != right_padding: - height_padding = top_padding + stride - width_padding = left_padding + stride - need_crop = True - padding = [height_padding, width_padding] - elif padding_type == "VALID": - height_padding = 0 - width_padding = 0 - padding = [height_padding, width_padding] - elif padding_type == "DYNAMIC": - padding = get_padding(filter_size, stride) - else: - padding = padding - - conv = fluid.layers.conv2d( - input, - num_filters, - filter_size, - groups=groups, - name=name, - stride=stride, - padding=padding, - use_cudnn=use_cudnn, - param_attr=param_attr, - bias_attr=bias_attr) - - if need_crop: - conv = conv[:, :, 1:, 1:] - - if norm is not None: - conv = norm_layer(input=conv, norm_type=norm, name=name + "_norm") - if act == 'relu': - conv = fluid.layers.relu(conv, name=name + '_relu') - elif act == 'leaky_relu': - conv = fluid.layers.leaky_relu( - conv, alpha=relufactor, name=name + '_leaky_relu') - elif act == 'tanh': - conv = fluid.layers.tanh(conv, name=name + '_tanh') - elif act == 'sigmoid': - conv = fluid.layers.sigmoid(conv, name=name + '_sigmoid') - elif act == 'swish': - conv = fluid.layers.swish(conv, name=name + '_swish') - elif act == None: - conv = conv - else: - raise NotImplementedError("activation: [%s] is not support" % act) - - return conv diff --git a/hub_module/modules/image/classification/efficientnetb1_imagenet/module.py b/hub_module/modules/image/classification/efficientnetb1_imagenet/module.py index cd46ec4c..70a0ac10 100644 --- a/hub_module/modules/image/classification/efficientnetb1_imagenet/module.py +++ b/hub_module/modules/image/classification/efficientnetb1_imagenet/module.py @@ -1,4 +1,3 @@ -# -*- coding:utf-8 -*- # copyright (c) 2020 PaddlePaddle Authors. All Rights Reserve. # # Licensed under the Apache License, Version 2.0 (the "License"); @@ -13,321 +12,747 @@ # See the License for the specific language governing permissions and # limitations under the License. -from __future__ import absolute_import -from __future__ import division - -import ast -import argparse import os - -import numpy as np -import paddle.fluid as fluid -import paddlehub as hub -from paddle.fluid.core import PaddleTensor, AnalysisConfig, create_paddle_predictor -from paddlehub.module.module import moduleinfo, runnable, serving -from paddlehub.common.paddle_helper import add_vars_prefix - -from efficientnetb1_imagenet.processor import postprocess, base64_to_cv2 -from efficientnetb1_imagenet.data_feed import reader -from efficientnetb1_imagenet.efficientnet import EfficientNetB1 - - -@moduleinfo( - name="efficientnetb1_imagenet", - type="CV/image_classification", - author="paddlepaddle", - author_email="paddle-dev@baidu.com", - summary= - "EfficientNetB1 is a image classfication model, this module is trained with imagenet datasets.", - version="1.1.0") -class EfficientNetB1ImageNet(hub.Module): - def _initialize(self): - self.default_pretrained_model_path = os.path.join( - self.directory, "efficientnetb1_imagenet_infer_model") - label_file = os.path.join(self.directory, "label_list.txt") - with open(label_file, 'r', encoding='utf-8') as file: - self.label_list = file.read().split("\n")[:-1] - - # It's compatibale with the usage of module's old_version. - self.classification = self.classify - - self._set_config() - - def get_expected_image_width(self): - return 224 - - def get_expected_image_height(self): - return 224 - - def get_pretrained_images_mean(self): - im_mean = np.array([0.485, 0.456, 0.406]).reshape(1, 3) - return im_mean - - def get_pretrained_images_std(self): - im_std = np.array([0.229, 0.224, 0.225]).reshape(1, 3) - return im_std - - def _set_config(self): - """ - predictor config setting - """ - cpu_config = AnalysisConfig(self.default_pretrained_model_path) - cpu_config.disable_glog_info() - cpu_config.disable_gpu() - self.cpu_predictor = create_paddle_predictor(cpu_config) - - try: - _places = os.environ["CUDA_VISIBLE_DEVICES"] - int(_places[0]) - use_gpu = True - except: - use_gpu = False - if use_gpu: - gpu_config = AnalysisConfig(self.default_pretrained_model_path) - gpu_config.disable_glog_info() - gpu_config.enable_use_gpu( - memory_pool_init_size_mb=1000, device_id=0) - self.gpu_predictor = create_paddle_predictor(gpu_config) - - def context(self, - trainable=True, - pretrained=True, - override_params=None, - phase='train'): - """context for transfer learning. - - Args: - trainable (bool): Set parameters in program to be trainable. - pretrained (bool) : Whether to load pretrained model. - - Returns: - inputs (dict): key is 'image', corresponding vaule is image tensor. - outputs (dict): key is : - 'classification', corresponding value is the result of classification. - 'feature_map', corresponding value is the result of the layer before the fully connected layer. - context_prog (fluid.Program): program for transfer learning. - """ - if phase in ["dev", "test", "predict", "eval"]: - is_test = True - elif phase in ["train"]: - is_test = False - else: - raise ValueError( - "Phase %s is error, which must be one of train, dev, test, eval and predict." - % phase) - - context_prog = fluid.Program() - startup_prog = fluid.Program() - with fluid.program_guard(context_prog, startup_prog): - with fluid.unique_name.guard(): - image = fluid.layers.data( - name="image", shape=[3, 224, 224], dtype="float32") - efficientnet_b1 = EfficientNetB1( - override_params=override_params) - output, feature_map = efficientnet_b1.net( - input=image, - class_dim=len(self.label_list), - is_test=is_test) - - name_prefix = '@HUB_{}@'.format(self.name) - inputs = {'image': name_prefix + image.name} - outputs = { - 'classification': name_prefix + output.name, - 'feature_map': name_prefix + feature_map.name - } - add_vars_prefix(context_prog, name_prefix) - add_vars_prefix(startup_prog, name_prefix) - global_vars = context_prog.global_block().vars - inputs = { - key: global_vars[value] - for key, value in inputs.items() - } - outputs = { - key: global_vars[value] - for key, value in outputs.items() - } - - place = fluid.CPUPlace() - exe = fluid.Executor(place) - # pretrained - if pretrained: - - def _if_exist(var): - b = os.path.exists( - os.path.join(self.default_pretrained_model_path, - var.name)) - return b - - fluid.io.load_vars( - exe, - self.default_pretrained_model_path, - context_prog, - predicate=_if_exist) - else: - exe.run(startup_prog) - # trainable - for param in context_prog.global_block().iter_parameters(): - param.trainable = trainable - return inputs, outputs, context_prog - - def classify(self, - images=None, - paths=None, - batch_size=1, - use_gpu=False, - top_k=1): +import math +import collections +import re +import copy + +import paddle +from paddle import ParamAttr +import paddle.nn as nn +import paddle.nn.functional as F +from paddle.nn import Conv2d, BatchNorm, Linear, Dropout +from paddle.nn import AdaptiveAvgPool2d, MaxPool2d, AvgPool2d +from paddlehub.module.module import moduleinfo +from paddlehub.module.cv_module import ImageClassifierModule + +GlobalParams = collections.namedtuple('GlobalParams', [ + 'batch_norm_momentum', + 'batch_norm_epsilon', + 'dropout_rate', + 'num_classes', + 'width_coefficient', + 'depth_coefficient', + 'depth_divisor', + 'min_depth', + 'drop_connect_rate', +]) + +BlockArgs = collections.namedtuple( + 'BlockArgs', + ['kernel_size', 'num_repeat', 'input_filters', 'output_filters', 'expand_ratio', 'id_skip', 'stride', 'se_ratio']) + +GlobalParams.__new__.__defaults__ = (None, ) * len(GlobalParams._fields) +BlockArgs.__new__.__defaults__ = (None, ) * len(BlockArgs._fields) + + +def efficientnet_params(model_name: str): + """ Map EfficientNet model name to parameter coefficients. """ + params_dict = { + # Coefficients: width,depth,resolution,dropout + 'efficientnet-b1': (1.0, 1.1, 240, 0.2) + } + return params_dict[model_name] + + +def efficientnet(width_coefficient: float = None, + depth_coefficient: float = None, + dropout_rate: float = 0.2, + drop_connect_rate: float = 0.2): + """ Get block arguments according to parameter and coefficients. """ + blocks_args = [ + 'r1_k3_s11_e1_i32_o16_se0.25', + 'r2_k3_s22_e6_i16_o24_se0.25', + 'r2_k5_s22_e6_i24_o40_se0.25', + 'r3_k3_s22_e6_i40_o80_se0.25', + 'r3_k5_s11_e6_i80_o112_se0.25', + 'r4_k5_s22_e6_i112_o192_se0.25', + 'r1_k3_s11_e6_i192_o320_se0.25', + ] + blocks_args = BlockDecoder.decode(blocks_args) + + global_params = GlobalParams(batch_norm_momentum=0.99, + batch_norm_epsilon=1e-3, + dropout_rate=dropout_rate, + drop_connect_rate=drop_connect_rate, + num_classes=1000, + width_coefficient=width_coefficient, + depth_coefficient=depth_coefficient, + depth_divisor=8, + min_depth=None) + + return blocks_args, global_params + + +def get_model_params(model_name: str, override_params: dict): + """ Get the block args and global params for a given model """ + if model_name.startswith('efficientnet'): + w, d, _, p = efficientnet_params(model_name) + blocks_args, global_params = efficientnet(width_coefficient=w, depth_coefficient=d, dropout_rate=p) + else: + raise NotImplementedError('model name is not pre-defined: %s' % model_name) + if override_params: + global_params = global_params._replace(**override_params) + return blocks_args, global_params + + +def round_filters(filters: int, global_params: dict): + """ Calculate and round number of filters based on depth multiplier. """ + multiplier = global_params.width_coefficient + if not multiplier: + return filters + divisor = global_params.depth_divisor + min_depth = global_params.min_depth + filters *= multiplier + min_depth = min_depth or divisor + new_filters = max(min_depth, int(filters + divisor / 2) // divisor * divisor) + if new_filters < 0.9 * filters: # prevent rounding by more than 10% + new_filters += divisor + return int(new_filters) + + +def round_repeats(repeats: int, global_params: dict): + """ Round number of filters based on depth multiplier. """ + multiplier = global_params.depth_coefficient + if not multiplier: + return repeats + return int(math.ceil(multiplier * repeats)) + + +class BlockDecoder(object): + """ + Block Decoder, straight from the official TensorFlow repository. + """ + @staticmethod + def _decode_block_string(block_string: str): + """ Gets a block through a string notation of arguments. """ + assert isinstance(block_string, str) + + ops = block_string.split('_') + options = {} + for op in ops: + splits = re.split(r'(\d.*)', op) + if len(splits) >= 2: + key, value = splits[:2] + options[key] = value + + # Check stride + cond_1 = ('s' in options and len(options['s']) == 1) + cond_2 = ((len(options['s']) == 2) and (options['s'][0] == options['s'][1])) + assert (cond_1 or cond_2) + + return BlockArgs(kernel_size=int(options['k']), + num_repeat=int(options['r']), + input_filters=int(options['i']), + output_filters=int(options['o']), + expand_ratio=int(options['e']), + id_skip=('noskip' not in block_string), + se_ratio=float(options['se']) if 'se' in options else None, + stride=[int(options['s'][0])]) + + @staticmethod + def _encode_block_string(block): + """Encodes a block to a string.""" + args = [ + 'r%d' % block.num_repeat, + 'k%d' % block.kernel_size, + 's%d%d' % (block.strides[0], block.strides[1]), + 'e%s' % block.expand_ratio, + 'i%d' % block.input_filters, + 'o%d' % block.output_filters + ] + if 0 < block.se_ratio <= 1: + args.append('se%s' % block.se_ratio) + if block.id_skip is False: + args.append('noskip') + return '_'.join(args) + + @staticmethod + def decode(string_list: list): """ - API for image classification. - - Args: - images (list[numpy.ndarray]): data of images, shape of each is [H, W, C], color space must be BGR. - paths (list[str]): The paths of images. - batch_size (int): batch size. - use_gpu (bool): Whether to use gpu. - top_k (int): Return top k results. + Decode a list of string notations to specify blocks in the network. - Returns: - res (list[dict]): The classfication results. + string_list: list of strings, each string is a notation of block + return + list of BlockArgs namedtuples of block args """ - if use_gpu: - try: - _places = os.environ["CUDA_VISIBLE_DEVICES"] - int(_places[0]) - except: - raise RuntimeError( - "Environment Variable CUDA_VISIBLE_DEVICES is not set correctly. If you wanna use gpu, please set CUDA_VISIBLE_DEVICES as cuda_device_id." - ) - - all_data = list() - for yield_data in reader(images, paths): - all_data.append(yield_data) - - total_num = len(all_data) - loop_num = int(np.ceil(total_num / batch_size)) - - res = list() - for iter_id in range(loop_num): - batch_data = list() - handle_id = iter_id * batch_size - for image_id in range(batch_size): - try: - batch_data.append(all_data[handle_id + image_id]) - except: - pass - # feed batch image - batch_image = np.array([data['image'] for data in batch_data]) - batch_image = PaddleTensor(batch_image.copy()) - predictor_output = self.gpu_predictor.run([ - batch_image - ]) if use_gpu else self.cpu_predictor.run([batch_image]) - out = postprocess( - data_out=predictor_output[0].as_ndarray(), - label_list=self.label_list, - top_k=top_k) - res += out - return res - - def save_inference_model(self, - dirname, - model_filename=None, - params_filename=None, - combined=True): - if combined: - model_filename = "__model__" if not model_filename else model_filename - params_filename = "__params__" if not params_filename else params_filename - place = fluid.CPUPlace() - exe = fluid.Executor(place) - - program, feeded_var_names, target_vars = fluid.io.load_inference_model( - dirname=self.default_pretrained_model_path, executor=exe) - - fluid.io.save_inference_model( - dirname=dirname, - main_program=program, - executor=exe, - feeded_var_names=feeded_var_names, - target_vars=target_vars, - model_filename=model_filename, - params_filename=params_filename) - - @serving - def serving_method(self, images, **kwargs): + assert isinstance(string_list, list) + blocks_args = [] + for block_string in string_list: + blocks_args.append(BlockDecoder._decode_block_string(block_string)) + return blocks_args + + @staticmethod + def encode(blocks_args: list): """ - Run as a service. - """ - images_decode = [base64_to_cv2(image) for image in images] - results = self.classify(images=images_decode, **kwargs) - return results + Encodes a list of BlockArgs to a list of strings. - @runnable - def run_cmd(self, argvs): - """ - Run as a command. + :param blocks_args: a list of BlockArgs namedtuples of block args + :return: a list of strings, each string is a notation of block """ - self.parser = argparse.ArgumentParser( - description="Run the {} module.".format(self.name), - prog='hub run {}'.format(self.name), - usage='%(prog)s', - add_help=True) - self.arg_input_group = self.parser.add_argument_group( - title="Input options", description="Input data. Required") - self.arg_config_group = self.parser.add_argument_group( - title="Config options", - description= - "Run configuration for controlling module behavior, not required.") - self.add_module_config_arg() - self.add_module_input_arg() - args = self.parser.parse_args(argvs) - results = self.classify( - paths=[args.input_path], - batch_size=args.batch_size, - use_gpu=args.use_gpu) - return results - - def add_module_config_arg(self): - """ - Add the command config options. - """ - self.arg_config_group.add_argument( - '--use_gpu', - type=ast.literal_eval, - default=False, - help="whether use GPU or not.") - self.arg_config_group.add_argument( - '--batch_size', - type=ast.literal_eval, - default=1, - help="batch size.") - self.arg_config_group.add_argument( - '--top_k', - type=ast.literal_eval, - default=1, - help="Return top k results.") - - def add_module_input_arg(self): - """ - Add the command input options. - """ - self.arg_input_group.add_argument( - '--input_path', type=str, help="path to image.") - - -if __name__ == '__main__': - b1 = EfficientNetB1ImageNet() - b1.context() - import cv2 - test_image = [ - cv2.imread( - '/mnt/zhangxuefei/program-paddle/PaddleHub/hub_module/tests/image_dataset/classification/animals/dog.jpeg' - ) - ] - res = b1.classification(images=test_image) - print(res) - res = b1.classification(paths=[ - '/mnt/zhangxuefei/program-paddle/PaddleHub/hub_module/tests/image_dataset/classification/animals/dog.jpeg' - ]) - print(res) - res = b1.classification(images=test_image) - print(res) - res = b1.classify(images=test_image) - print(res) + block_strings = [] + for block in blocks_args: + block_strings.append(BlockDecoder._encode_block_string(block)) + return block_strings + + +def initial_type(name: str, use_bias: bool = False): + param_attr = ParamAttr(name=name + "_weights") + if use_bias: + bias_attr = ParamAttr(name=name + "_offset") + else: + bias_attr = False + return param_attr, bias_attr + + +def init_batch_norm_layer(name: str = "batch_norm"): + param_attr = ParamAttr(name=name + "_scale") + bias_attr = ParamAttr(name=name + "_offset") + return param_attr, bias_attr + + +def init_fc_layer(name: str = "fc"): + param_attr = ParamAttr(name=name + "_weights") + bias_attr = ParamAttr(name=name + "_offset") + return param_attr, bias_attr + + +def cal_padding(img_size: int, stride: int, filter_size: int, dilation: int = 1): + """Calculate padding size.""" + if img_size % stride == 0: + out_size = max(filter_size - stride, 0) + else: + out_size = max(filter_size - (img_size % stride), 0) + return out_size // 2, out_size - out_size // 2 + + +inp_shape = {"b1": [240, 120, 120, 60, 30, 15, 15, 8]} + + +def _drop_connect(inputs: paddle.Tensor, prob: float, is_test: bool): + """Drop input connection""" + if is_test: + return inputs + keep_prob = 1.0 - prob + inputs_shape = paddle.shape(inputs) + random_tensor = keep_prob + paddle.rand(shape=[inputs_shape[0], 1, 1, 1]) + binary_tensor = paddle.floor(random_tensor) + output = inputs / keep_prob * binary_tensor + return output + + +class Conv2ds(nn.Layer): + """Basic conv layer""" + def __init__(self, + input_channels: int, + output_channels: int, + filter_size: int, + stride: int = 1, + padding: int = 0, + groups: int = None, + name: str = "conv2d", + act: str = None, + use_bias: bool = False, + padding_type: str = None, + model_name: str = None, + cur_stage: str = None): + super(Conv2ds, self).__init__() + assert act in [None, "swish", "sigmoid"] + self.act = act + + param_attr, bias_attr = initial_type(name=name, use_bias=use_bias) + + def get_padding(filter_size, stride=1, dilation=1): + padding = ((stride - 1) + dilation * (filter_size - 1)) // 2 + return padding + + inps = 1 if model_name == None and cur_stage == None else inp_shape[model_name][cur_stage] + self.need_crop = False + if padding_type == "SAME": + top_padding, bottom_padding = cal_padding(inps, stride, filter_size) + left_padding, right_padding = cal_padding(inps, stride, filter_size) + height_padding = bottom_padding + width_padding = right_padding + if top_padding != bottom_padding or left_padding != right_padding: + height_padding = top_padding + stride + width_padding = left_padding + stride + self.need_crop = True + padding = [height_padding, width_padding] + elif padding_type == "VALID": + height_padding = 0 + width_padding = 0 + padding = [height_padding, width_padding] + elif padding_type == "DYNAMIC": + padding = get_padding(filter_size, stride) + else: + padding = padding_type + + groups = 1 if groups is None else groups + self._conv = Conv2d(input_channels, + output_channels, + filter_size, + groups=groups, + stride=stride, + padding=padding, + weight_attr=param_attr, + bias_attr=bias_attr) + + def forward(self, inputs: paddle.Tensor): + x = self._conv(inputs) + if self.act == "swish": + x = F.swish(x) + elif self.act == "sigmoid": + x = F.sigmoid(x) + + if self.need_crop: + x = x[:, :, 1:, 1:] + return x + + +class ConvBNLayer(nn.Layer): + """Basic conv bn layer.""" + def __init__(self, + input_channels: int, + filter_size: int, + output_channels: int, + stride: int = 1, + num_groups: int = 1, + padding_type: str = "SAME", + conv_act: str = None, + bn_act: str = "swish", + use_bn: bool = True, + use_bias: bool = False, + name: str = None, + conv_name: str = None, + bn_name: str = None, + model_name: str = None, + cur_stage: str = None): + super(ConvBNLayer, self).__init__() + + self._conv = Conv2ds(input_channels=input_channels, + output_channels=output_channels, + filter_size=filter_size, + stride=stride, + groups=num_groups, + act=conv_act, + padding_type=padding_type, + name=conv_name, + use_bias=use_bias, + model_name=model_name, + cur_stage=cur_stage) + self.use_bn = use_bn + if use_bn is True: + bn_name = name + bn_name + param_attr, bias_attr = init_batch_norm_layer(bn_name) + + self._bn = BatchNorm(num_channels=output_channels, + act=bn_act, + momentum=0.99, + epsilon=0.001, + moving_mean_name=bn_name + "_mean", + moving_variance_name=bn_name + "_variance", + param_attr=param_attr, + bias_attr=bias_attr) + + def forward(self, inputs: paddle.Tensor): + if self.use_bn: + x = self._conv(inputs) + x = self._bn(x) + return x + else: + return self._conv(inputs) + + +class ExpandConvNorm(nn.Layer): + """Expand conv norm layer.""" + def __init__(self, + input_channels: int, + block_args: dict, + padding_type: str, + name: str = None, + model_name: str = None, + cur_stage: str = None): + super(ExpandConvNorm, self).__init__() + + self.oup = block_args.input_filters * block_args.expand_ratio + self.expand_ratio = block_args.expand_ratio + + if self.expand_ratio != 1: + self._conv = ConvBNLayer(input_channels, + 1, + self.oup, + bn_act=None, + padding_type=padding_type, + name=name, + conv_name=name + "_expand_conv", + bn_name="_bn0", + model_name=model_name, + cur_stage=cur_stage) + + def forward(self, inputs: paddle.Tensor): + if self.expand_ratio != 1: + return self._conv(inputs) + else: + return inputs + + +class DepthwiseConvNorm(nn.Layer): + """Depthwise conv norm layer.""" + def __init__(self, + input_channels: int, + block_args: dict, + padding_type: str, + name: str = None, + model_name: str = None, + cur_stage: str = None): + super(DepthwiseConvNorm, self).__init__() + + self.k = block_args.kernel_size + self.s = block_args.stride + if isinstance(self.s, list) or isinstance(self.s, tuple): + self.s = self.s[0] + oup = block_args.input_filters * block_args.expand_ratio + + self._conv = ConvBNLayer(input_channels, + self.k, + oup, + self.s, + num_groups=input_channels, + bn_act=None, + padding_type=padding_type, + name=name, + conv_name=name + "_depthwise_conv", + bn_name="_bn1", + model_name=model_name, + cur_stage=cur_stage) + + def forward(self, inputs: paddle.Tensor): + return self._conv(inputs) + + +class ProjectConvNorm(nn.Layer): + """Projection conv bn layer.""" + def __init__(self, + input_channels: int, + block_args: dict, + padding_type: str, + name: str = None, + model_name: str = None, + cur_stage: str = None): + super(ProjectConvNorm, self).__init__() + + final_oup = block_args.output_filters + + self._conv = ConvBNLayer(input_channels, + 1, + final_oup, + bn_act=None, + padding_type=padding_type, + name=name, + conv_name=name + "_project_conv", + bn_name="_bn2", + model_name=model_name, + cur_stage=cur_stage) + + def forward(self, inputs: paddle.Tensor): + return self._conv(inputs) + + +class SEBlock(nn.Layer): + """Basic Squeeze-and-Excitation block for Efficientnet.""" + def __init__(self, + input_channels: int, + num_squeezed_channels: int, + oup: int, + padding_type: str, + name: str = None, + model_name: str = None, + cur_stage: str = None): + super(SEBlock, self).__init__() + + self._pool = AdaptiveAvgPool2d(1) + self._conv1 = Conv2ds(input_channels, + num_squeezed_channels, + 1, + use_bias=True, + padding_type=padding_type, + act="swish", + name=name + "_se_reduce") + + self._conv2 = Conv2ds(num_squeezed_channels, + oup, + 1, + act="sigmoid", + use_bias=True, + padding_type=padding_type, + name=name + "_se_expand") + + def forward(self, inputs: paddle.Tensor): + x = self._pool(inputs) + x = self._conv1(x) + x = self._conv2(x) + return paddle.multiply(inputs, x) + + +class MbConvBlock(nn.Layer): + """Mobile inverted bottleneck convolution for Efficientnet.""" + def __init__(self, + input_channels: int, + block_args: dict, + padding_type: str, + use_se: bool, + name: str = None, + drop_connect_rate: float = None, + is_test: bool = False, + model_name: str = None, + cur_stage: str = None): + super(MbConvBlock, self).__init__() + + oup = block_args.input_filters * block_args.expand_ratio + self.block_args = block_args + self.has_se = use_se and (block_args.se_ratio is not None) and (0 < block_args.se_ratio <= 1) + self.id_skip = block_args.id_skip + self.expand_ratio = block_args.expand_ratio + self.drop_connect_rate = drop_connect_rate + self.is_test = is_test + + if self.expand_ratio != 1: + self._ecn = ExpandConvNorm(input_channels, + block_args, + padding_type=padding_type, + name=name, + model_name=model_name, + cur_stage=cur_stage) + + self._dcn = DepthwiseConvNorm(input_channels * block_args.expand_ratio, + block_args, + padding_type=padding_type, + name=name, + model_name=model_name, + cur_stage=cur_stage) + + if self.has_se: + num_squeezed_channels = max(1, int(block_args.input_filters * block_args.se_ratio)) + self._se = SEBlock(input_channels * block_args.expand_ratio, + num_squeezed_channels, + oup, + padding_type=padding_type, + name=name, + model_name=model_name, + cur_stage=cur_stage) + + self._pcn = ProjectConvNorm(input_channels * block_args.expand_ratio, + block_args, + padding_type=padding_type, + name=name, + model_name=model_name, + cur_stage=cur_stage) + + def forward(self, inputs: paddle.Tensor): + x = inputs + if self.expand_ratio != 1: + x = self._ecn(x) + x = F.swish(x) + x = self._dcn(x) + x = F.swish(x) + if self.has_se: + x = self._se(x) + x = self._pcn(x) + if self.id_skip and \ + self.block_args.stride == 1 and \ + self.block_args.input_filters == self.block_args.output_filters: + if self.drop_connect_rate: + x = _drop_connect(x, self.drop_connect_rate, self.is_test) + x = paddle.elementwise_add(x, inputs) + return x + + +class ConvStemNorm(nn.Layer): + """Basic conv stem norm block for extracting features.""" + def __init__(self, + input_channels: int, + padding_type: str, + _global_params: dict, + name: str = None, + model_name: str = None, + cur_stage: str = None): + super(ConvStemNorm, self).__init__() + + output_channels = round_filters(32, _global_params) + self._conv = ConvBNLayer(input_channels, + filter_size=3, + output_channels=output_channels, + stride=2, + bn_act=None, + padding_type=padding_type, + name="", + conv_name="_conv_stem", + bn_name="_bn0", + model_name=model_name, + cur_stage=cur_stage) + + def forward(self, inputs: paddle.Tensor): + return self._conv(inputs) + + +class ExtractFeatures(nn.Layer): + """Extract features.""" + def __init__(self, + input_channels: int, + _block_args: dict, + _global_params: dict, + padding_type: str, + use_se: bool, + is_test: bool, + model_name: str = None): + super(ExtractFeatures, self).__init__() + + self._global_params = _global_params + + self._conv_stem = ConvStemNorm(input_channels, + padding_type=padding_type, + _global_params=_global_params, + model_name=model_name, + cur_stage=0) + + self.block_args_copy = copy.deepcopy(_block_args) + idx = 0 + block_size = 0 + for block_arg in self.block_args_copy: + block_arg = block_arg._replace(input_filters=round_filters(block_arg.input_filters, _global_params), + output_filters=round_filters(block_arg.output_filters, _global_params), + num_repeat=round_repeats(block_arg.num_repeat, _global_params)) + block_size += 1 + for _ in range(block_arg.num_repeat - 1): + block_size += 1 + + self.conv_seq = [] + cur_stage = 1 + for block_args in _block_args: + block_args = block_args._replace(input_filters=round_filters(block_args.input_filters, _global_params), + output_filters=round_filters(block_args.output_filters, _global_params), + num_repeat=round_repeats(block_args.num_repeat, _global_params)) + + drop_connect_rate = self._global_params.drop_connect_rate if not is_test else 0 + if drop_connect_rate: + drop_connect_rate *= float(idx) / block_size + + _mc_block = self.add_sublayer( + "_blocks." + str(idx) + ".", + MbConvBlock(block_args.input_filters, + block_args=block_args, + padding_type=padding_type, + use_se=use_se, + name="_blocks." + str(idx) + ".", + drop_connect_rate=drop_connect_rate, + model_name=model_name, + cur_stage=cur_stage)) + self.conv_seq.append(_mc_block) + idx += 1 + if block_args.num_repeat > 1: + block_args = block_args._replace(input_filters=block_args.output_filters, stride=1) + for _ in range(block_args.num_repeat - 1): + drop_connect_rate = self._global_params.drop_connect_rate if not is_test else 0 + if drop_connect_rate: + drop_connect_rate *= float(idx) / block_size + _mc_block = self.add_sublayer( + "block." + str(idx) + ".", + MbConvBlock(block_args.input_filters, + block_args, + padding_type=padding_type, + use_se=use_se, + name="_blocks." + str(idx) + ".", + drop_connect_rate=drop_connect_rate, + model_name=model_name, + cur_stage=cur_stage)) + self.conv_seq.append(_mc_block) + idx += 1 + cur_stage += 1 + + def forward(self, inputs: paddle.Tensor): + x = self._conv_stem(inputs) + x = F.swish(x) + for _mc_block in self.conv_seq: + x = _mc_block(x) + return x + + +@moduleinfo(name="efficientnetb1_imagenet", + type="cv/classification", + author="paddlepaddle", + author_email="", + summary="efficientnetb1_imagenet is a classification model, " + "this module is trained with Imagenet dataset.", + version="1.1.0", + meta=ImageClassifierModule) +class EfficientNet_B1(nn.Layer): + def __init__(self, + is_test: bool = False, + padding_type: str = "SAME", + override_params: dict = None, + use_se: bool = True, + class_dim: int = 1000, + load_checkpoint: str = None): + super(EfficientNet_B1, self).__init__() + + model_name = 'efficientnet-b1' + self.name = "b1" + self._block_args, self._global_params = get_model_params(model_name, override_params) + self.padding_type = padding_type + self.use_se = use_se + self.is_test = is_test + + self._ef = ExtractFeatures(3, + self._block_args, + self._global_params, + self.padding_type, + self.use_se, + self.is_test, + model_name=self.name) + + output_channels = round_filters(1280, self._global_params) + oup = 320 + + self._conv = ConvBNLayer(oup, + 1, + output_channels, + bn_act="swish", + padding_type=self.padding_type, + name="", + conv_name="_conv_head", + bn_name="_bn1", + model_name=self.name, + cur_stage=7) + self._pool = AdaptiveAvgPool2d(1) + + if self._global_params.dropout_rate: + self._drop = Dropout(p=self._global_params.dropout_rate, mode="upscale_in_train") + + param_attr, bias_attr = init_fc_layer("_fc") + self._fc = Linear(output_channels, class_dim, weight_attr=param_attr, bias_attr=bias_attr) + + if load_checkpoint is not None: + model_dict = paddle.load(load_checkpoint)[0] + self.set_dict(model_dict) + print("load custom checkpoint success") + + else: + checkpoint = os.path.join(self.directory, 'efficientnet_b1_imagenet.pdparams') + if not os.path.exists(checkpoint): + os.system( + 'wget https://paddlehub.bj.bcebos.com/dygraph/image_classification/efficientnet_b1_imagenet.pdparams -O ' + + checkpoint) + model_dict = paddle.load(checkpoint)[0] + self.set_dict(model_dict) + print("load pretrained checkpoint success") + + def forward(self, inputs: paddle.Tensor): + x = self._ef(inputs) + x = self._conv(x) + x = self._pool(x) + if self._global_params.dropout_rate: + x = self._drop(x) + x = paddle.squeeze(x, axis=[2, 3]) + x = self._fc(x) + return x diff --git a/hub_module/modules/image/classification/efficientnetb1_imagenet/processor.py b/hub_module/modules/image/classification/efficientnetb1_imagenet/processor.py deleted file mode 100644 index 12ccc9b2..00000000 --- a/hub_module/modules/image/classification/efficientnetb1_imagenet/processor.py +++ /dev/null @@ -1,69 +0,0 @@ -# -*- coding:utf-8 -*- -# copyright (c) 2020 PaddlePaddle Authors. All Rights Reserve. -# -# Licensed under the Apache License, Version 2.0 (the "License"); -# you may not use this file except in compliance with the License. -# You may obtain a copy of the License at -# -# http://www.apache.org/licenses/LICENSE-2.0 -# -# Unless required by applicable law or agreed to in writing, software -# distributed under the License is distributed on an "AS IS" BASIS, -# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. -# See the License for the specific language governing permissions and -# limitations under the License. - -from __future__ import absolute_import -from __future__ import division -from __future__ import print_function - -import base64 -import cv2 -import os - -import numpy as np - - -def base64_to_cv2(b64str): - data = base64.b64decode(b64str.encode('utf8')) - data = np.fromstring(data, np.uint8) - data = cv2.imdecode(data, cv2.IMREAD_COLOR) - return data - - -def softmax(x): - orig_shape = x.shape - if len(x.shape) > 1: - tmp = np.max(x, axis=1) - x -= tmp.reshape((x.shape[0], 1)) - x = np.exp(x) - tmp = np.sum(x, axis=1) - x /= tmp.reshape((x.shape[0], 1)) - else: - tmp = np.max(x) - x -= tmp - x = np.exp(x) - tmp = np.sum(x) - x /= tmp - return x - - -def postprocess(data_out, label_list, top_k): - """ - Postprocess output of network, one image at a time. - - Args: - data_out (numpy.ndarray): output data of network. - label_list (list): list of label. - top_k (int): Return top k results. - """ - output = [] - for result in data_out: - result_i = softmax(result) - output_i = {} - indexs = np.argsort(result_i)[::-1][0:top_k] - for index in indexs: - label = label_list[index].split(',')[0] - output_i[label] = float(result_i[index]) - output.append(output_i) - return output diff --git a/hub_module/modules/image/classification/efficientnetb2_imagenet/README.md b/hub_module/modules/image/classification/efficientnetb2_imagenet/README.md deleted file mode 100644 index eda1a225..00000000 --- a/hub_module/modules/image/classification/efficientnetb2_imagenet/README.md +++ /dev/null @@ -1,149 +0,0 @@ -## 命令行预测 - -``` -hub run efficientnetb2_imagenet --input_path "/PATH/TO/IMAGE" -``` - -## API - -```python -def get_expected_image_width() -``` - -返回预处理的图片宽度，也就是224。 - -```python -def get_expected_image_height() -``` - -返回预处理的图片高度，也就是224。 - -```python -def get_pretrained_images_mean() -``` - -返回预处理的图片均值，也就是 \[0.485, 0.456, 0.406\]。 - -```python -def get_pretrained_images_std() -``` - -返回预处理的图片标准差，也就是 \[0.229, 0.224, 0.225\]。 - - -```python -def context(trainable=True, pretrained=True) -``` - -**参数** - -* trainable (bool): 计算图的参数是否为可训练的； -* pretrained (bool): 是否加载默认的预训练模型。 - -**返回** - -* inputs (dict): 计算图的输入，key 为 'image', value 为图片的张量； -* outputs (dict): 计算图的输出，key 为 'classification' 和 'feature_map'，其相应的值为： - * classification (paddle.fluid.framework.Variable): 分类结果，也就是全连接层的输出； - * feature\_map (paddle.fluid.framework.Variable): 特征匹配，全连接层前面的那个张量。 -* context\_prog(fluid.Program): 计算图，用于迁移学习。 - -```python -def classify(images=None, - paths=None, - batch_size=1, - use_gpu=False, - top_k=1): -``` - -**参数** - -* images (list\[numpy.ndarray\]): 图片数据，每一个图片数据的shape 均为 \[H, W, C\]，颜色空间为 BGR； -* paths (list\[str\]): 图片的路径； -* batch\_size (int): batch 的大小； -* use\_gpu (bool): 是否使用 GPU 来预测； -* top\_k (int): 返回预测结果的前 k 个。 - -**返回** - -res (list\[dict\]): 分类结果，列表的每一个元素均为字典，其中 key 为识别动物的类别，value为置信度。 - -```python -def save_inference_model(dirname, - model_filename=None, - params_filename=None, - combined=True) -``` - -将模型保存到指定路径。 - -**参数** - -* dirname: 存在模型的目录名称 -* model\_filename: 模型文件名称，默认为\_\_model\_\_ -* params\_filename: 参数文件名称，默认为\_\_params\_\_(仅当`combined`为True时生效) -* combined: 是否将参数保存到统一的一个文件中 - -## 代码示例 - -```python -import paddlehub as hub -import cv2 - -classifier = hub.Module(name="efficientnetb2_imagenet") - -result = classifier.classify(images=[cv2.imread('/PATH/TO/IMAGE')]) -# or -# result = classifier.classify(paths=['/PATH/TO/IMAGE']) -``` - -## 服务部署 - -PaddleHub Serving可以部署一个在线图像识别服务。 - -## 第一步：启动PaddleHub Serving - -运行启动命令： -```shell -$ hub serving start -m efficientnetb2_imagenet -``` - -这样就完成了一个在线图像识别服务化API的部署，默认端口号为8866。 - -**NOTE:** 如使用GPU预测，则需要在启动服务之前，请设置CUDA\_VISIBLE\_DEVICES环境变量，否则不用设置。 - -## 第二步：发送预测请求 - -配置好服务端，以下数行代码即可实现发送预测请求，获取预测结果 - -```python -import requests -import json -import cv2 -import base64 - - -def cv2_to_base64(image): - data = cv2.imencode('.jpg', image)[1] - return base64.b64encode(data.tostring()).decode('utf8') - - -# 发送HTTP请求 -data = {'images':[cv2_to_base64(cv2.imread("/PATH/TO/IMAGE"))]} -headers = {"Content-type": "application/json"} -url = "http://127.0.0.1:8866/predict/efficientnetb2_imagenet" -r = requests.post(url=url, headers=headers, data=json.dumps(data)) - -# 打印预测结果 -print(r.json()["results"]) -``` - -### 查看代码 - -https://github.com/PaddlePaddle/PaddleClas - -### 依赖 - -paddlepaddle >= 1.6.2 - -paddlehub >= 1.6.0 diff --git a/hub_module/modules/image/classification/efficientnetb2_imagenet/__init__.py b/hub_module/modules/image/classification/efficientnetb2_imagenet/__init__.py deleted file mode 100644 index 19a990da..00000000 --- a/hub_module/modules/image/classification/efficientnetb2_imagenet/__init__.py +++ /dev/null @@ -1 +0,0 @@ -# -*- coding:utf-8 -*- diff --git a/hub_module/modules/image/classification/efficientnetb2_imagenet/data_feed.py b/hub_module/modules/image/classification/efficientnetb2_imagenet/data_feed.py deleted file mode 100644 index 11181458..00000000 --- a/hub_module/modules/image/classification/efficientnetb2_imagenet/data_feed.py +++ /dev/null @@ -1,98 +0,0 @@ -# -*- coding:utf-8 -*- -# copyright (c) 2020 PaddlePaddle Authors. All Rights Reserve. -# -# Licensed under the Apache License, Version 2.0 (the "License"); -# you may not use this file except in compliance with the License. -# You may obtain a copy of the License at -# -# http://www.apache.org/licenses/LICENSE-2.0 -# -# Unless required by applicable law or agreed to in writing, software -# distributed under the License is distributed on an "AS IS" BASIS, -# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. -# See the License for the specific language governing permissions and -# limitations under the License. - -import os -import time -from collections import OrderedDict - -import cv2 -import numpy as np -from PIL import Image - -__all__ = ['reader'] - -DATA_DIM = 224 -img_mean = np.array([0.485, 0.456, 0.406]).reshape((3, 1, 1)) -img_std = np.array([0.229, 0.224, 0.225]).reshape((3, 1, 1)) - - -def resize_short(img, target_size): - percent = float(target_size) / min(img.size[0], img.size[1]) - resized_width = int(round(img.size[0] * percent)) - resized_height = int(round(img.size[1] * percent)) - img = img.resize((resized_width, resized_height), Image.LANCZOS) - return img - - -def crop_image(img, target_size, center): - width, height = img.size - size = target_size - if center == True: - w_start = (width - size) / 2 - h_start = (height - size) / 2 - else: - w_start = np.random.randint(0, width - size + 1) - h_start = np.random.randint(0, height - size + 1) - w_end = w_start + size - h_end = h_start + size - img = img.crop((w_start, h_start, w_end, h_end)) - return img - - -def process_image(img): - img = resize_short(img, target_size=256) - img = crop_image(img, target_size=DATA_DIM, center=True) - if img.mode != 'RGB': - img = img.convert('RGB') - img = np.array(img).astype('float32').transpose((2, 0, 1)) / 255 - img -= img_mean - img /= img_std - return img - - -def reader(images=None, paths=None): - """ - Preprocess to yield image. - - Args: - images (list[numpy.ndarray]): images data, shape of each is [H, W, C]. - paths (list[str]): paths to images. - - Yield: - each (collections.OrderedDict): info of original image, preprocessed image. - """ - component = list() - if paths: - for im_path in paths: - each = OrderedDict() - assert os.path.isfile( - im_path), "The {} isn't a valid file path.".format(im_path) - each['org_im_path'] = im_path - each['org_im'] = Image.open(im_path) - each['org_im_width'], each['org_im_height'] = each['org_im'].size - component.append(each) - if images is not None: - assert type(images), "images is a list." - for im in images: - each = OrderedDict() - each['org_im'] = Image.fromarray(im[:, :, ::-1]) - each['org_im_path'] = 'ndarray_time={}'.format( - round(time.time(), 6) * 1e6) - each['org_im_width'], each['org_im_height'] = each['org_im'].size - component.append(each) - - for element in component: - element['image'] = process_image(element['org_im']) - yield element diff --git a/hub_module/modules/image/classification/efficientnetb2_imagenet/efficientnet.py b/hub_module/modules/image/classification/efficientnetb2_imagenet/efficientnet.py deleted file mode 100644 index 6a9d08c2..00000000 --- a/hub_module/modules/image/classification/efficientnetb2_imagenet/efficientnet.py +++ /dev/null @@ -1,623 +0,0 @@ -# -*- coding:utf-8 -*- -# copyright (c) 2020 PaddlePaddle Authors. All Rights Reserve. -# -# Licensed under the Apache License, Version 2.0 (the "License"); -# you may not use this file except in compliance with the License. -# You may obtain a copy of the License at -# -# http://www.apache.org/licenses/LICENSE-2.0 -# -# Unless required by applicable law or agreed to in writing, software -# distributed under the License is distributed on an "AS IS" BASIS, -# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. -# See the License for the specific language governing permissions and -# limitations under the License. - -from __future__ import absolute_import -from __future__ import division -from __future__ import print_function - -import collections -import re -import math -import copy - -import paddle.fluid as fluid -from efficientnetb2_imagenet.layers import conv2d, init_batch_norm_layer, init_fc_layer - -__all__ = [ - 'EfficientNet', 'EfficientNetB0_small', 'EfficientNetB0', 'EfficientNetB1', - 'EfficientNetB2', 'EfficientNetB3', 'EfficientNetB4', 'EfficientNetB5', - 'EfficientNetB6', 'EfficientNetB7' -] - -GlobalParams = collections.namedtuple('GlobalParams', [ - 'batch_norm_momentum', - 'batch_norm_epsilon', - 'dropout_rate', - 'num_classes', - 'width_coefficient', - 'depth_coefficient', - 'depth_divisor', - 'min_depth', - 'drop_connect_rate', -]) - -BlockArgs = collections.namedtuple('BlockArgs', [ - 'kernel_size', 'num_repeat', 'input_filters', 'output_filters', - 'expand_ratio', 'id_skip', 'stride', 'se_ratio' -]) - -GlobalParams.__new__.__defaults__ = (None, ) * len(GlobalParams._fields) -BlockArgs.__new__.__defaults__ = (None, ) * len(BlockArgs._fields) - - -def efficientnet_params(model_name): - """ Map EfficientNet model name to parameter coefficients. """ - params_dict = { - # Coefficients: width,depth,resolution,dropout - 'efficientnet-b0': (1.0, 1.0, 224, 0.2), - 'efficientnet-b1': (1.0, 1.1, 240, 0.2), - 'efficientnet-b2': (1.1, 1.2, 260, 0.3), - 'efficientnet-b3': (1.2, 1.4, 300, 0.3), - 'efficientnet-b4': (1.4, 1.8, 380, 0.4), - 'efficientnet-b5': (1.6, 2.2, 456, 0.4), - 'efficientnet-b6': (1.8, 2.6, 528, 0.5), - 'efficientnet-b7': (2.0, 3.1, 600, 0.5), - } - return params_dict[model_name] - - -def efficientnet(width_coefficient=None, - depth_coefficient=None, - dropout_rate=0.2, - drop_connect_rate=0.2): - """ Get block arguments according to parameter and coefficients. """ - blocks_args = [ - 'r1_k3_s11_e1_i32_o16_se0.25', - 'r2_k3_s22_e6_i16_o24_se0.25', - 'r2_k5_s22_e6_i24_o40_se0.25', - 'r3_k3_s22_e6_i40_o80_se0.25', - 'r3_k5_s11_e6_i80_o112_se0.25', - 'r4_k5_s22_e6_i112_o192_se0.25', - 'r1_k3_s11_e6_i192_o320_se0.25', - ] - blocks_args = BlockDecoder.decode(blocks_args) - - global_params = GlobalParams( - batch_norm_momentum=0.99, - batch_norm_epsilon=1e-3, - dropout_rate=dropout_rate, - drop_connect_rate=drop_connect_rate, - num_classes=1000, - width_coefficient=width_coefficient, - depth_coefficient=depth_coefficient, - depth_divisor=8, - min_depth=None) - - return blocks_args, global_params - - -def get_model_params(model_name, override_params): - """ Get the block args and global params for a given model """ - if model_name.startswith('efficientnet'): - w, d, _, p = efficientnet_params(model_name) - blocks_args, global_params = efficientnet( - width_coefficient=w, depth_coefficient=d, dropout_rate=p) - else: - raise NotImplementedError( - 'model name is not pre-defined: %s' % model_name) - if override_params: - global_params = global_params._replace(**override_params) - return blocks_args, global_params - - -def round_filters(filters, global_params): - """ Calculate and round number of filters based on depth multiplier. """ - multiplier = global_params.width_coefficient - if not multiplier: - return filters - divisor = global_params.depth_divisor - min_depth = global_params.min_depth - filters *= multiplier - min_depth = min_depth or divisor - new_filters = max(min_depth, - int(filters + divisor / 2) // divisor * divisor) - if new_filters < 0.9 * filters: # prevent rounding by more than 10% - new_filters += divisor - return int(new_filters) - - -def round_repeats(repeats, global_params): - """ Round number of filters based on depth multiplier. """ - multiplier = global_params.depth_coefficient - if not multiplier: - return repeats - return int(math.ceil(multiplier * repeats)) - - -class EfficientNet(): - def __init__(self, - name='b0', - padding_type='SAME', - override_params=None, - is_test=False, - use_se=True): - valid_names = ['b' + str(i) for i in range(8)] - assert name in valid_names, 'efficient name should be in b0~b7' - model_name = 'efficientnet-' + name - self._blocks_args, self._global_params = get_model_params( - model_name, override_params) - self._bn_mom = self._global_params.batch_norm_momentum - self._bn_eps = self._global_params.batch_norm_epsilon - self.padding_type = padding_type - self.use_se = use_se - - def net(self, input, class_dim=1000, is_test=False): - - conv = self.extract_features(input, is_test=is_test) - - out_channels = round_filters(1280, self._global_params) - conv = self.conv_bn_layer( - conv, - num_filters=out_channels, - filter_size=1, - bn_act='swish', - bn_mom=self._bn_mom, - bn_eps=self._bn_eps, - padding_type=self.padding_type, - name='', - conv_name='_conv_head', - bn_name='_bn1') - - pool = fluid.layers.pool2d( - input=conv, pool_type='avg', global_pooling=True, use_cudnn=False) - - if not is_test and self._global_params.dropout_rate: - pool = fluid.layers.dropout( - pool, - self._global_params.dropout_rate, - dropout_implementation='upscale_in_train') - - param_attr, bias_attr = init_fc_layer(class_dim, '_fc') - out = fluid.layers.fc( - pool, - class_dim, - name='_fc', - param_attr=param_attr, - bias_attr=bias_attr) - return out, pool - - def _drop_connect(self, inputs, prob, is_test): - if is_test: - return inputs - keep_prob = 1.0 - prob - random_tensor = keep_prob + fluid.layers.uniform_random_batch_size_like( - inputs, [-1, 1, 1, 1], min=0., max=1.) - binary_tensor = fluid.layers.floor(random_tensor) - output = inputs / keep_prob * binary_tensor - return output - - def _expand_conv_norm(self, inputs, block_args, is_test, name=None): - # Expansion phase - oup = block_args.input_filters * block_args.expand_ratio # number of output channels - - if block_args.expand_ratio != 1: - conv = self.conv_bn_layer( - inputs, - num_filters=oup, - filter_size=1, - bn_act=None, - bn_mom=self._bn_mom, - bn_eps=self._bn_eps, - padding_type=self.padding_type, - name=name, - conv_name=name + '_expand_conv', - bn_name='_bn0') - - return conv - - def _depthwise_conv_norm(self, inputs, block_args, is_test, name=None): - k = block_args.kernel_size - s = block_args.stride - if isinstance(s, list) or isinstance(s, tuple): - s = s[0] - oup = block_args.input_filters * block_args.expand_ratio # number of output channels - - conv = self.conv_bn_layer( - inputs, - num_filters=oup, - filter_size=k, - stride=s, - num_groups=oup, - bn_act=None, - padding_type=self.padding_type, - bn_mom=self._bn_mom, - bn_eps=self._bn_eps, - name=name, - use_cudnn=False, - conv_name=name + '_depthwise_conv', - bn_name='_bn1') - - return conv - - def _project_conv_norm(self, inputs, block_args, is_test, name=None): - final_oup = block_args.output_filters - conv = self.conv_bn_layer( - inputs, - num_filters=final_oup, - filter_size=1, - bn_act=None, - padding_type=self.padding_type, - bn_mom=self._bn_mom, - bn_eps=self._bn_eps, - name=name, - conv_name=name + '_project_conv', - bn_name='_bn2') - return conv - - def conv_bn_layer(self, - input, - filter_size, - num_filters, - stride=1, - num_groups=1, - padding_type="SAME", - conv_act=None, - bn_act='swish', - use_cudnn=True, - use_bn=True, - bn_mom=0.9, - bn_eps=1e-05, - use_bias=False, - name=None, - conv_name=None, - bn_name=None): - conv = conv2d( - input=input, - num_filters=num_filters, - filter_size=filter_size, - stride=stride, - groups=num_groups, - act=conv_act, - padding_type=padding_type, - use_cudnn=use_cudnn, - name=conv_name, - use_bias=use_bias) - - if use_bn == False: - return conv - else: - bn_name = name + bn_name - param_attr, bias_attr = init_batch_norm_layer(bn_name) - return fluid.layers.batch_norm( - input=conv, - act=bn_act, - momentum=bn_mom, - epsilon=bn_eps, - name=bn_name, - moving_mean_name=bn_name + '_mean', - moving_variance_name=bn_name + '_variance', - param_attr=param_attr, - bias_attr=bias_attr) - - def _conv_stem_norm(self, inputs, is_test): - out_channels = round_filters(32, self._global_params) - bn = self.conv_bn_layer( - inputs, - num_filters=out_channels, - filter_size=3, - stride=2, - bn_act=None, - bn_mom=self._bn_mom, - padding_type=self.padding_type, - bn_eps=self._bn_eps, - name='', - conv_name='_conv_stem', - bn_name='_bn0') - - return bn - - def mb_conv_block(self, - inputs, - block_args, - is_test=False, - drop_connect_rate=None, - name=None): - # Expansion and Depthwise Convolution - oup = block_args.input_filters * block_args.expand_ratio # number of output channels - has_se = self.use_se and (block_args.se_ratio is - not None) and (0 < block_args.se_ratio <= 1) - id_skip = block_args.id_skip # skip connection and drop connect - conv = inputs - if block_args.expand_ratio != 1: - conv = fluid.layers.swish( - self._expand_conv_norm(conv, block_args, is_test, name)) - - conv = fluid.layers.swish( - self._depthwise_conv_norm(conv, block_args, is_test, name)) - - # Squeeze and Excitation - if has_se: - num_squeezed_channels = max( - 1, int(block_args.input_filters * block_args.se_ratio)) - conv = self.se_block(conv, num_squeezed_channels, oup, name) - - conv = self._project_conv_norm(conv, block_args, is_test, name) - - # Skip connection and drop connect - input_filters, output_filters = block_args.input_filters, block_args.output_filters - if id_skip and block_args.stride == 1 and input_filters == output_filters: - if drop_connect_rate: - conv = self._drop_connect(conv, drop_connect_rate, is_test) - conv = fluid.layers.elementwise_add(conv, inputs) - - return conv - - def se_block(self, inputs, num_squeezed_channels, oup, name): - x_squeezed = fluid.layers.pool2d( - input=inputs, pool_type='avg', global_pooling=True, use_cudnn=False) - x_squeezed = conv2d( - x_squeezed, - num_filters=num_squeezed_channels, - filter_size=1, - use_bias=True, - padding_type=self.padding_type, - act='swish', - name=name + '_se_reduce') - x_squeezed = conv2d( - x_squeezed, - num_filters=oup, - filter_size=1, - use_bias=True, - padding_type=self.padding_type, - name=name + '_se_expand') - se_out = inputs * fluid.layers.sigmoid(x_squeezed) - return se_out - - def extract_features(self, inputs, is_test): - """ Returns output of the final convolution layer """ - - conv = fluid.layers.swish(self._conv_stem_norm(inputs, is_test=is_test)) - - block_args_copy = copy.deepcopy(self._blocks_args) - idx = 0 - block_size = 0 - for block_arg in block_args_copy: - block_arg = block_arg._replace( - input_filters=round_filters(block_arg.input_filters, - self._global_params), - output_filters=round_filters(block_arg.output_filters, - self._global_params), - num_repeat=round_repeats(block_arg.num_repeat, - self._global_params)) - block_size += 1 - for _ in range(block_arg.num_repeat - 1): - block_size += 1 - - for block_args in self._blocks_args: - - # Update block input and output filters based on depth multiplier. - block_args = block_args._replace( - input_filters=round_filters(block_args.input_filters, - self._global_params), - output_filters=round_filters(block_args.output_filters, - self._global_params), - num_repeat=round_repeats(block_args.num_repeat, - self._global_params)) - - # The first block needs to take care of stride and filter size increase. - drop_connect_rate = self._global_params.drop_connect_rate - if drop_connect_rate: - drop_connect_rate *= float(idx) / block_size - conv = self.mb_conv_block(conv, block_args, is_test, - drop_connect_rate, - '_blocks.' + str(idx) + '.') - - idx += 1 - if block_args.num_repeat > 1: - block_args = block_args._replace( - input_filters=block_args.output_filters, stride=1) - for _ in range(block_args.num_repeat - 1): - drop_connect_rate = self._global_params.drop_connect_rate - if drop_connect_rate: - drop_connect_rate *= float(idx) / block_size - conv = self.mb_conv_block(conv, block_args, is_test, - drop_connect_rate, - '_blocks.' + str(idx) + '.') - idx += 1 - - return conv - - def shortcut(self, input, data_residual): - return fluid.layers.elementwise_add(input, data_residual) - - -class BlockDecoder(object): - """ Block Decoder for readability, straight from the official TensorFlow repository """ - - @staticmethod - def _decode_block_string(block_string): - """ Gets a block through a string notation of arguments. """ - assert isinstance(block_string, str) - - ops = block_string.split('_') - options = {} - for op in ops: - splits = re.split(r'(\d.*)', op) - if len(splits) >= 2: - key, value = splits[:2] - options[key] = value - - # Check stride - assert (('s' in options and len(options['s']) == 1) or - (len(options['s']) == 2 and options['s'][0] == options['s'][1])) - - return BlockArgs( - kernel_size=int(options['k']), - num_repeat=int(options['r']), - input_filters=int(options['i']), - output_filters=int(options['o']), - expand_ratio=int(options['e']), - id_skip=('noskip' not in block_string), - se_ratio=float(options['se']) if 'se' in options else None, - stride=[int(options['s'][0])]) - - @staticmethod - def _encode_block_string(block): - """Encodes a block to a string.""" - args = [ - 'r%d' % block.num_repeat, - 'k%d' % block.kernel_size, - 's%d%d' % (block.strides[0], block.strides[1]), - 'e%s' % block.expand_ratio, - 'i%d' % block.input_filters, - 'o%d' % block.output_filters - ] - if 0 < block.se_ratio <= 1: - args.append('se%s' % block.se_ratio) - if block.id_skip is False: - args.append('noskip') - return '_'.join(args) - - @staticmethod - def decode(string_list): - """ - Decodes a list of string notations to specify blocks inside the network. - :param string_list: a list of strings, each string is a notation of block - :return: a list of BlockArgs namedtuples of block args - """ - assert isinstance(string_list, list) - blocks_args = [] - for block_string in string_list: - blocks_args.append(BlockDecoder._decode_block_string(block_string)) - return blocks_args - - @staticmethod - def encode(blocks_args): - """ - Encodes a list of BlockArgs to a list of strings. - :param blocks_args: a list of BlockArgs namedtuples of block args - :return: a list of strings, each string is a notation of block - """ - block_strings = [] - for block in blocks_args: - block_strings.append(BlockDecoder._encode_block_string(block)) - return block_strings - - -def EfficientNetB0_small(is_test=False, - padding_type='SAME', - override_params=None, - use_se=False): - model = EfficientNet( - name='b0', - is_test=is_test, - padding_type=padding_type, - override_params=override_params, - use_se=use_se) - return model - - -def EfficientNetB0(is_test=False, - padding_type='SAME', - override_params=None, - use_se=True): - model = EfficientNet( - name='b0', - is_test=is_test, - padding_type=padding_type, - override_params=override_params, - use_se=use_se) - return model - - -def EfficientNetB1(is_test=False, - padding_type='SAME', - override_params=None, - use_se=True): - model = EfficientNet( - name='b1', - is_test=is_test, - padding_type=padding_type, - override_params=override_params, - use_se=use_se) - return model - - -def EfficientNetB2(is_test=False, - padding_type='SAME', - override_params=None, - use_se=True): - model = EfficientNet( - name='b2', - is_test=is_test, - padding_type=padding_type, - override_params=override_params, - use_se=use_se) - return model - - -def EfficientNetB3(is_test=False, - padding_type='SAME', - override_params=None, - use_se=True): - model = EfficientNet( - name='b3', - is_test=is_test, - padding_type=padding_type, - override_params=override_params, - use_se=use_se) - return model - - -def EfficientNetB4(is_test=False, - padding_type='SAME', - override_params=None, - use_se=True): - model = EfficientNet( - name='b4', - is_test=is_test, - padding_type=padding_type, - override_params=override_params, - use_se=use_se) - return model - - -def EfficientNetB5(is_test=False, - padding_type='SAME', - override_params=None, - use_se=True): - model = EfficientNet( - name='b5', - is_test=is_test, - padding_type=padding_type, - override_params=override_params, - use_se=use_se) - return model - - -def EfficientNetB6(is_test=False, - padding_type='SAME', - override_params=None, - use_se=True): - model = EfficientNet( - name='b6', - is_test=is_test, - padding_type=padding_type, - override_params=override_params, - use_se=use_se) - return model - - -def EfficientNetB7(is_test=False, - padding_type='SAME', - override_params=None, - use_se=True): - model = EfficientNet( - name='b7', - is_test=is_test, - padding_type=padding_type, - override_params=override_params, - use_se=use_se) - return model diff --git a/hub_module/modules/image/classification/efficientnetb2_imagenet/label_list.txt b/hub_module/modules/image/classification/efficientnetb2_imagenet/label_list.txt deleted file mode 100644 index a509c007..00000000 --- a/hub_module/modules/image/classification/efficientnetb2_imagenet/label_list.txt +++ /dev/null @@ -1,1000 +0,0 @@ -tench, Tinca tinca -goldfish, Carassius auratus -great white shark, white shark, man-eater, man-eating shark, Carcharodon carcharias -tiger shark, Galeocerdo cuvieri -hammerhead, hammerhead shark -electric ray, crampfish, numbfish, torpedo -stingray -cock -hen -ostrich, Struthio camelus -brambling, Fringilla montifringilla -goldfinch, Carduelis carduelis -house finch, linnet, Carpodacus mexicanus -junco, snowbird -indigo bunting, indigo finch, indigo bird, Passerina cyanea -robin, American robin, Turdus migratorius -bulbul -jay -magpie -chickadee -water ouzel, dipper -kite -bald eagle, American eagle, Haliaeetus leucocephalus -vulture -great grey owl, great gray owl, Strix nebulosa -European fire salamander, Salamandra salamandra -common newt, Triturus vulgaris -eft -spotted salamander, Ambystoma maculatum -axolotl, mud puppy, Ambystoma mexicanum -bullfrog, Rana catesbeiana -tree frog, tree-frog -tailed frog, bell toad, ribbed toad, tailed toad, Ascaphus trui -loggerhead, loggerhead turtle, Caretta caretta -leatherback turtle, leatherback, leathery turtle, Dermochelys coriacea -mud turtle -terrapin -box turtle, box tortoise -banded gecko -common iguana, iguana, Iguana iguana -American chameleon, anole, Anolis carolinensis -whiptail, whiptail lizard -agama -frilled lizard, Chlamydosaurus kingi -alligator lizard -Gila monster, Heloderma suspectum -green lizard, Lacerta viridis -African chameleon, Chamaeleo chamaeleon -Komodo dragon, Komodo lizard, dragon lizard, giant lizard, Varanus komodoensis -African crocodile, Nile crocodile, Crocodylus niloticus -American alligator, Alligator mississipiensis -triceratops -thunder snake, worm snake, Carphophis amoenus -ringneck snake, ring-necked snake, ring snake -hognose snake, puff adder, sand viper -green snake, grass snake -king snake, kingsnake -garter snake, grass snake -water snake -vine snake -night snake, Hypsiglena torquata -boa constrictor, Constrictor constrictor -rock python, rock snake, Python sebae -Indian cobra, Naja naja -green mamba -sea snake -horned viper, cerastes, sand viper, horned asp, Cerastes cornutus -diamondback, diamondback rattlesnake, Crotalus adamanteus -sidewinder, horned rattlesnake, Crotalus cerastes -trilobite -harvestman, daddy longlegs, Phalangium opilio -scorpion -black and gold garden spider, Argiope aurantia -barn spider, Araneus cavaticus -garden spider, Aranea diademata -black widow, Latrodectus mactans -tarantula -wolf spider, hunting spider -tick -centipede -black grouse -ptarmigan -ruffed grouse, partridge, Bonasa umbellus -prairie chicken, prairie grouse, prairie fowl -peacock -quail -partridge -African grey, African gray, Psittacus erithacus -macaw -sulphur-crested cockatoo, Kakatoe galerita, Cacatua galerita -lorikeet -coucal -bee eater -hornbill -hummingbird -jacamar -toucan -drake -red-breasted merganser, Mergus serrator -goose -black swan, Cygnus atratus -tusker -echidna, spiny anteater, anteater -platypus, duckbill, duckbilled platypus, duck-billed platypus, Ornithorhynchus anatinus -wallaby, brush kangaroo -koala, koala bear, kangaroo bear, native bear, Phascolarctos cinereus -wombat -jellyfish -sea anemone, anemone -brain coral -flatworm, platyhelminth -nematode, nematode worm, roundworm -conch -snail -slug -sea slug, nudibranch -chiton, coat-of-mail shell, sea cradle, polyplacophore -chambered nautilus, pearly nautilus, nautilus -Dungeness crab, Cancer magister -rock crab, Cancer irroratus -fiddler crab -king crab, Alaska crab, Alaskan king crab, Alaska king crab, Paralithodes camtschatica -American lobster, Northern lobster, Maine lobster, Homarus americanus -spiny lobster, langouste, rock lobster, crawfish, crayfish, sea crawfish -crayfish, crawfish, crawdad, crawdaddy -hermit crab -isopod -white stork, Ciconia ciconia -black stork, Ciconia nigra -spoonbill -flamingo -little blue heron, Egretta caerulea -American egret, great white heron, Egretta albus -bittern -crane -limpkin, Aramus pictus -European gallinule, Porphyrio porphyrio -American coot, marsh hen, mud hen, water hen, Fulica americana -bustard -ruddy turnstone, Arenaria interpres -red-backed sandpiper, dunlin, Erolia alpina -redshank, Tringa totanus -dowitcher -oystercatcher, oyster catcher -pelican -king penguin, Aptenodytes patagonica -albatross, mollymawk -grey whale, gray whale, devilfish, Eschrichtius gibbosus, Eschrichtius robustus -killer whale, killer, orca, grampus, sea wolf, Orcinus orca -dugong, Dugong dugon -sea lion -Chihuahua -Japanese spaniel -Maltese dog, Maltese terrier, Maltese -Pekinese, Pekingese, Peke -Shih-Tzu -Blenheim spaniel -papillon -toy terrier -Rhodesian ridgeback -Afghan hound, Afghan -basset, basset hound -beagle -bloodhound, sleuthhound -bluetick -black-and-tan coonhound -Walker hound, Walker foxhound -English foxhound -redbone -borzoi, Russian wolfhound -Irish wolfhound -Italian greyhound -whippet -Ibizan hound, Ibizan Podenco -Norwegian elkhound, elkhound -otterhound, otter hound -Saluki, gazelle hound -Scottish deerhound, deerhound -Weimaraner -Staffordshire bullterrier, Staffordshire bull terrier -American Staffordshire terrier, Staffordshire terrier, American pit bull terrier, pit bull terrier -Bedlington terrier -Border terrier -Kerry blue terrier -Irish terrier -Norfolk terrier -Norwich terrier -Yorkshire terrier -wire-haired fox terrier -Lakeland terrier -Sealyham terrier, Sealyham -Airedale, Airedale terrier -cairn, cairn terrier -Australian terrier -Dandie Dinmont, Dandie Dinmont terrier -Boston bull, Boston terrier -miniature schnauzer -giant schnauzer -standard schnauzer -Scotch terrier, Scottish terrier, Scottie -Tibetan terrier, chrysanthemum dog -silky terrier, Sydney silky -soft-coated wheaten terrier -West Highland white terrier -Lhasa, Lhasa apso -flat-coated retriever -curly-coated retriever -golden retriever -Labrador retriever -Chesapeake Bay retriever -German short-haired pointer -vizsla, Hungarian pointer -English setter -Irish setter, red setter -Gordon setter -Brittany spaniel -clumber, clumber spaniel -English springer, English springer spaniel -Welsh springer spaniel -cocker spaniel, English cocker spaniel, cocker -Sussex spaniel -Irish water spaniel -kuvasz -schipperke -groenendael -malinois -briard -kelpie -komondor -Old English sheepdog, bobtail -Shetland sheepdog, Shetland sheep dog, Shetland -collie -Border collie -Bouvier des Flandres, Bouviers des Flandres -Rottweiler -German shepherd, German shepherd dog, German police dog, alsatian -Doberman, Doberman pinscher -miniature pinscher -Greater Swiss Mountain dog -Bernese mountain dog -Appenzeller -EntleBucher -boxer -bull mastiff -Tibetan mastiff -French bulldog -Great Dane -Saint Bernard, St Bernard -Eskimo dog, husky -malamute, malemute, Alaskan malamute -Siberian husky -dalmatian, coach dog, carriage dog -affenpinscher, monkey pinscher, monkey dog -basenji -pug, pug-dog -Leonberg -Newfoundland, Newfoundland dog -Great Pyrenees -Samoyed, Samoyede -Pomeranian -chow, chow chow -keeshond -Brabancon griffon -Pembroke, Pembroke Welsh corgi -Cardigan, Cardigan Welsh corgi -toy poodle -miniature poodle -standard poodle -Mexican hairless -timber wolf, grey wolf, gray wolf, Canis lupus -white wolf, Arctic wolf, Canis lupus tundrarum -red wolf, maned wolf, Canis rufus, Canis niger -coyote, prairie wolf, brush wolf, Canis latrans -dingo, warrigal, warragal, Canis dingo -dhole, Cuon alpinus -African hunting dog, hyena dog, Cape hunting dog, Lycaon pictus -hyena, hyaena -red fox, Vulpes vulpes -kit fox, Vulpes macrotis -Arctic fox, white fox, Alopex lagopus -grey fox, gray fox, Urocyon cinereoargenteus -tabby, tabby cat -tiger cat -Persian cat -Siamese cat, Siamese -Egyptian cat -cougar, puma, catamount, mountain lion, painter, panther, Felis concolor -lynx, catamount -leopard, Panthera pardus -snow leopard, ounce, Panthera uncia -jaguar, panther, Panthera onca, Felis onca -lion, king of beasts, Panthera leo -tiger, Panthera tigris -cheetah, chetah, Acinonyx jubatus -brown bear, bruin, Ursus arctos -American black bear, black bear, Ursus americanus, Euarctos americanus -ice bear, polar bear, Ursus Maritimus, Thalarctos maritimus -sloth bear, Melursus ursinus, Ursus ursinus -mongoose -meerkat, mierkat -tiger beetle -ladybug, ladybeetle, lady beetle, ladybird, ladybird beetle -ground beetle, carabid beetle -long-horned beetle, longicorn, longicorn beetle -leaf beetle, chrysomelid -dung beetle -rhinoceros beetle -weevil -fly -bee -ant, emmet, pismire -grasshopper, hopper -cricket -walking stick, walkingstick, stick insect -cockroach, roach -mantis, mantid -cicada, cicala -leafhopper -lacewing, lacewing fly -dragonfly, darning needle, devil's darning needle, sewing needle, snake feeder, snake doctor, mosquito hawk, skeeter hawk -damselfly -admiral -ringlet, ringlet butterfly -monarch, monarch butterfly, milkweed butterfly, Danaus plexippus -cabbage butterfly -sulphur butterfly, sulfur butterfly -lycaenid, lycaenid butterfly -starfish, sea star -sea urchin -sea cucumber, holothurian -wood rabbit, cottontail, cottontail rabbit -hare -Angora, Angora rabbit -hamster -porcupine, hedgehog -fox squirrel, eastern fox squirrel, Sciurus niger -marmot -beaver -guinea pig, Cavia cobaya -sorrel -zebra -hog, pig, grunter, squealer, Sus scrofa -wild boar, boar, Sus scrofa -warthog -hippopotamus, hippo, river horse, Hippopotamus amphibius -ox -water buffalo, water ox, Asiatic buffalo, Bubalus bubalis -bison -ram, tup -bighorn, bighorn sheep, cimarron, Rocky Mountain bighorn, Rocky Mountain sheep, Ovis canadensis -ibex, Capra ibex -hartebeest -impala, Aepyceros melampus -gazelle -Arabian camel, dromedary, Camelus dromedarius -llama -weasel -mink -polecat, fitch, foulmart, foumart, Mustela putorius -black-footed ferret, ferret, Mustela nigripes -otter -skunk, polecat, wood pussy -badger -armadillo -three-toed sloth, ai, Bradypus tridactylus -orangutan, orang, orangutang, Pongo pygmaeus -gorilla, Gorilla gorilla -chimpanzee, chimp, Pan troglodytes -gibbon, Hylobates lar -siamang, Hylobates syndactylus, Symphalangus syndactylus -guenon, guenon monkey -patas, hussar monkey, Erythrocebus patas -baboon -macaque -langur -colobus, colobus monkey -proboscis monkey, Nasalis larvatus -marmoset -capuchin, ringtail, Cebus capucinus -howler monkey, howler -titi, titi monkey -spider monkey, Ateles geoffroyi -squirrel monkey, Saimiri sciureus -Madagascar cat, ring-tailed lemur, Lemur catta -indri, indris, Indri indri, Indri brevicaudatus -Indian elephant, Elephas maximus -African elephant, Loxodonta africana -lesser panda, red panda, panda, bear cat, cat bear, Ailurus fulgens -giant panda, panda, panda bear, coon bear, Ailuropoda melanoleuca -barracouta, snoek -eel -coho, cohoe, coho salmon, blue jack, silver salmon, Oncorhynchus kisutch -rock beauty, Holocanthus tricolor -anemone fish -sturgeon -gar, garfish, garpike, billfish, Lepisosteus osseus -lionfish -puffer, pufferfish, blowfish, globefish -abacus -abaya -academic gown, academic robe, judge's robe -accordion, piano accordion, squeeze box -acoustic guitar -aircraft carrier, carrier, flattop, attack aircraft carrier -airliner -airship, dirigible -altar -ambulance -amphibian, amphibious vehicle -analog clock -apiary, bee house -apron -ashcan, trash can, garbage can, wastebin, ash bin, ash-bin, ashbin, dustbin, trash barrel, trash bin -assault rifle, assault gun -backpack, back pack, knapsack, packsack, rucksack, haversack -bakery, bakeshop, bakehouse -balance beam, beam -balloon -ballpoint, ballpoint pen, ballpen, Biro -Band Aid -banjo -bannister, banister, balustrade, balusters, handrail -barbell -barber chair -barbershop -barn -barometer -barrel, cask -barrow, garden cart, lawn cart, wheelbarrow -baseball -basketball -bassinet -bassoon -bathing cap, swimming cap -bath towel -bathtub, bathing tub, bath, tub -beach wagon, station wagon, wagon, estate car, beach waggon, station waggon, waggon -beacon, lighthouse, beacon light, pharos -beaker -bearskin, busby, shako -beer bottle -beer glass -bell cote, bell cot -bib -bicycle-built-for-two, tandem bicycle, tandem -bikini, two-piece -binder, ring-binder -binoculars, field glasses, opera glasses -birdhouse -boathouse -bobsled, bobsleigh, bob -bolo tie, bolo, bola tie, bola -bonnet, poke bonnet -bookcase -bookshop, bookstore, bookstall -bottlecap -bow -bow tie, bow-tie, bowtie -brass, memorial tablet, plaque -brassiere, bra, bandeau -breakwater, groin, groyne, mole, bulwark, seawall, jetty -breastplate, aegis, egis -broom -bucket, pail -buckle -bulletproof vest -bullet train, bullet -butcher shop, meat market -cab, hack, taxi, taxicab -caldron, cauldron -candle, taper, wax light -cannon -canoe -can opener, tin opener -cardigan -car mirror -carousel, carrousel, merry-go-round, roundabout, whirligig -carpenter's kit, tool kit -carton -car wheel -cash machine, cash dispenser, automated teller machine, automatic teller machine, automated teller, automatic teller, ATM -cassette -cassette player -castle -catamaran -CD player -cello, violoncello -cellular telephone, cellular phone, cellphone, cell, mobile phone -chain -chainlink fence -chain mail, ring mail, mail, chain armor, chain armour, ring armor, ring armour -chain saw, chainsaw -chest -chiffonier, commode -chime, bell, gong -china cabinet, china closet -Christmas stocking -church, church building -cinema, movie theater, movie theatre, movie house, picture palace -cleaver, meat cleaver, chopper -cliff dwelling -cloak -clog, geta, patten, sabot -cocktail shaker -coffee mug -coffeepot -coil, spiral, volute, whorl, helix -combination lock -computer keyboard, keypad -confectionery, confectionary, candy store -container ship, containership, container vessel -convertible -corkscrew, bottle screw -cornet, horn, trumpet, trump -cowboy boot -cowboy hat, ten-gallon hat -cradle -crane -crash helmet -crate -crib, cot -Crock Pot -croquet ball -crutch -cuirass -dam, dike, dyke -desk -desktop computer -dial telephone, dial phone -diaper, nappy, napkin -digital clock -digital watch -dining table, board -dishrag, dishcloth -dishwasher, dish washer, dishwashing machine -disk brake, disc brake -dock, dockage, docking facility -dogsled, dog sled, dog sleigh -dome -doormat, welcome mat -drilling platform, offshore rig -drum, membranophone, tympan -drumstick -dumbbell -Dutch oven -electric fan, blower -electric guitar -electric locomotive -entertainment center -envelope -espresso maker -face powder -feather boa, boa -file, file cabinet, filing cabinet -fireboat -fire engine, fire truck -fire screen, fireguard -flagpole, flagstaff -flute, transverse flute -folding chair -football helmet -forklift -fountain -fountain pen -four-poster -freight car -French horn, horn -frying pan, frypan, skillet -fur coat -garbage truck, dustcart -gasmask, respirator, gas helmet -gas pump, gasoline pump, petrol pump, island dispenser -goblet -go-kart -golf ball -golfcart, golf cart -gondola -gong, tam-tam -gown -grand piano, grand -greenhouse, nursery, glasshouse -grille, radiator grille -grocery store, grocery, food market, market -guillotine -hair slide -hair spray -half track -hammer -hamper -hand blower, blow dryer, blow drier, hair dryer, hair drier -hand-held computer, hand-held microcomputer -handkerchief, hankie, hanky, hankey -hard disc, hard disk, fixed disk -harmonica, mouth organ, harp, mouth harp -harp -harvester, reaper -hatchet -holster -home theater, home theatre -honeycomb -hook, claw -hoopskirt, crinoline -horizontal bar, high bar -horse cart, horse-cart -hourglass -iPod -iron, smoothing iron -jack-o'-lantern -jean, blue jean, denim -jeep, landrover -jersey, T-shirt, tee shirt -jigsaw puzzle -jinrikisha, ricksha, rickshaw -joystick -kimono -knee pad -knot -lab coat, laboratory coat -ladle -lampshade, lamp shade -laptop, laptop computer -lawn mower, mower -lens cap, lens cover -letter opener, paper knife, paperknife -library -lifeboat -lighter, light, igniter, ignitor -limousine, limo -liner, ocean liner -lipstick, lip rouge -Loafer -lotion -loudspeaker, speaker, speaker unit, loudspeaker system, speaker system -loupe, jeweler's loupe -lumbermill, sawmill -magnetic compass -mailbag, postbag -mailbox, letter box -maillot -maillot, tank suit -manhole cover -maraca -marimba, xylophone -mask -matchstick -maypole -maze, labyrinth -measuring cup -medicine chest, medicine cabinet -megalith, megalithic structure -microphone, mike -microwave, microwave oven -military uniform -milk can -minibus -miniskirt, mini -minivan -missile -mitten -mixing bowl -mobile home, manufactured home -Model T -modem -monastery -monitor -moped -mortar -mortarboard -mosque -mosquito net -motor scooter, scooter -mountain bike, all-terrain bike, off-roader -mountain tent -mouse, computer mouse -mousetrap -moving van -muzzle -nail -neck brace -necklace -nipple -notebook, notebook computer -obelisk -oboe, hautboy, hautbois -ocarina, sweet potato -odometer, hodometer, mileometer, milometer -oil filter -organ, pipe organ -oscilloscope, scope, cathode-ray oscilloscope, CRO -overskirt -oxcart -oxygen mask -packet -paddle, boat paddle -paddlewheel, paddle wheel -padlock -paintbrush -pajama, pyjama, pj's, jammies -palace -panpipe, pandean pipe, syrinx -paper towel -parachute, chute -parallel bars, bars -park bench -parking meter -passenger car, coach, carriage -patio, terrace -pay-phone, pay-station -pedestal, plinth, footstall -pencil box, pencil case -pencil sharpener -perfume, essence -Petri dish -photocopier -pick, plectrum, plectron -pickelhaube -picket fence, paling -pickup, pickup truck -pier -piggy bank, penny bank -pill bottle -pillow -ping-pong ball -pinwheel -pirate, pirate ship -pitcher, ewer -plane, carpenter's plane, woodworking plane -planetarium -plastic bag -plate rack -plow, plough -plunger, plumber's helper -Polaroid camera, Polaroid Land camera -pole -police van, police wagon, paddy wagon, patrol wagon, wagon, black Maria -poncho -pool table, billiard table, snooker table -pop bottle, soda bottle -pot, flowerpot -potter's wheel -power drill -prayer rug, prayer mat -printer -prison, prison house -projectile, missile -projector -puck, hockey puck -punching bag, punch bag, punching ball, punchball -purse -quill, quill pen -quilt, comforter, comfort, puff -racer, race car, racing car -racket, racquet -radiator -radio, wireless -radio telescope, radio reflector -rain barrel -recreational vehicle, RV, R.V. -reel -reflex camera -refrigerator, icebox -remote control, remote -restaurant, eating house, eating place, eatery -revolver, six-gun, six-shooter -rifle -rocking chair, rocker -rotisserie -rubber eraser, rubber, pencil eraser -rugby ball -rule, ruler -running shoe -safe -safety pin -saltshaker, salt shaker -sandal -sarong -sax, saxophone -scabbard -scale, weighing machine -school bus -schooner -scoreboard -screen, CRT screen -screw -screwdriver -seat belt, seatbelt -sewing machine -shield, buckler -shoe shop, shoe-shop, shoe store -shoji -shopping basket -shopping cart -shovel -shower cap -shower curtain -ski -ski mask -sleeping bag -slide rule, slipstick -sliding door -slot, one-armed bandit -snorkel -snowmobile -snowplow, snowplough -soap dispenser -soccer ball -sock -solar dish, solar collector, solar furnace -sombrero -soup bowl -space bar -space heater -space shuttle -spatula -speedboat -spider web, spider's web -spindle -sports car, sport car -spotlight, spot -stage -steam locomotive -steel arch bridge -steel drum -stethoscope -stole -stone wall -stopwatch, stop watch -stove -strainer -streetcar, tram, tramcar, trolley, trolley car -stretcher -studio couch, day bed -stupa, tope -submarine, pigboat, sub, U-boat -suit, suit of clothes -sundial -sunglass -sunglasses, dark glasses, shades -sunscreen, sunblock, sun blocker -suspension bridge -swab, swob, mop -sweatshirt -swimming trunks, bathing trunks -swing -switch, electric switch, electrical switch -syringe -table lamp -tank, army tank, armored combat vehicle, armoured combat vehicle -tape player -teapot -teddy, teddy bear -television, television system -tennis ball -thatch, thatched roof -theater curtain, theatre curtain -thimble -thresher, thrasher, threshing machine -throne -tile roof -toaster -tobacco shop, tobacconist shop, tobacconist -toilet seat -torch -totem pole -tow truck, tow car, wrecker -toyshop -tractor -trailer truck, tractor trailer, trucking rig, rig, articulated lorry, semi -tray -trench coat -tricycle, trike, velocipede -trimaran -tripod -triumphal arch -trolleybus, trolley coach, trackless trolley -trombone -tub, vat -turnstile -typewriter keyboard -umbrella -unicycle, monocycle -upright, upright piano -vacuum, vacuum cleaner -vase -vault -velvet -vending machine -vestment -viaduct -violin, fiddle -volleyball -waffle iron -wall clock -wallet, billfold, notecase, pocketbook -wardrobe, closet, press -warplane, military plane -washbasin, handbasin, washbowl, lavabo, wash-hand basin -washer, automatic washer, washing machine -water bottle -water jug -water tower -whiskey jug -whistle -wig -window screen -window shade -Windsor tie -wine bottle -wing -wok -wooden spoon -wool, woolen, woollen -worm fence, snake fence, snake-rail fence, Virginia fence -wreck -yawl -yurt -web site, website, internet site, site -comic book -crossword puzzle, crossword -street sign -traffic light, traffic signal, stoplight -book jacket, dust cover, dust jacket, dust wrapper -menu -plate -guacamole -consomme -hot pot, hotpot -trifle -ice cream, icecream -ice lolly, lolly, lollipop, popsicle -French loaf -bagel, beigel -pretzel -cheeseburger -hotdog, hot dog, red hot -mashed potato -head cabbage -broccoli -cauliflower -zucchini, courgette -spaghetti squash -acorn squash -butternut squash -cucumber, cuke -artichoke, globe artichoke -bell pepper -cardoon -mushroom -Granny Smith -strawberry -orange -lemon -fig -pineapple, ananas -banana -jackfruit, jak, jack -custard apple -pomegranate -hay -carbonara -chocolate sauce, chocolate syrup -dough -meat loaf, meatloaf -pizza, pizza pie -potpie -burrito -red wine -espresso -cup -eggnog -alp -bubble -cliff, drop, drop-off -coral reef -geyser -lakeside, lakeshore -promontory, headland, head, foreland -sandbar, sand bar -seashore, coast, seacoast, sea-coast -valley, vale -volcano -ballplayer, baseball player -groom, bridegroom -scuba diver -rapeseed -daisy -yellow lady's slipper, yellow lady-slipper, Cypripedium calceolus, Cypripedium parviflorum -corn -acorn -hip, rose hip, rosehip -buckeye, horse chestnut, conker -coral fungus -agaric -gyromitra -stinkhorn, carrion fungus -earthstar -hen-of-the-woods, hen of the woods, Polyporus frondosus, Grifola frondosa -bolete -ear, spike, capitulum -toilet tissue, toilet paper, bathroom tissue diff --git a/hub_module/modules/image/classification/efficientnetb2_imagenet/layers.py b/hub_module/modules/image/classification/efficientnetb2_imagenet/layers.py deleted file mode 100644 index 00f792b5..00000000 --- a/hub_module/modules/image/classification/efficientnetb2_imagenet/layers.py +++ /dev/null @@ -1,248 +0,0 @@ -# -*- coding:utf-8 -*- -# copyright (c) 2020 PaddlePaddle Authors. All Rights Reserve. -# -# Licensed under the Apache License, Version 2.0 (the "License"); -# you may not use this file except in compliance with the License. -# You may obtain a copy of the License at -# -# http://www.apache.org/licenses/LICENSE-2.0 -# -# Unless required by applicable law or agreed to in writing, software -# distributed under the License is distributed on an "AS IS" BASIS, -# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. -# See the License for the specific language governing permissions and -# limitations under the License. - -from __future__ import absolute_import -from __future__ import division -from __future__ import print_function - -import math -import warnings - -import paddle.fluid as fluid - - -def initial_type(name, - input, - op_type, - fan_out, - init="google", - use_bias=False, - filter_size=0, - stddev=0.02): - if init == "kaiming": - if op_type == 'conv': - fan_in = input.shape[1] * filter_size * filter_size - elif op_type == 'deconv': - fan_in = fan_out * filter_size * filter_size - else: - if len(input.shape) > 2: - fan_in = input.shape[1] * input.shape[2] * input.shape[3] - else: - fan_in = input.shape[1] - bound = 1 / math.sqrt(fan_in) - param_attr = fluid.ParamAttr( - name=name + "_weights", - initializer=fluid.initializer.Uniform(low=-bound, high=bound)) - if use_bias == True: - bias_attr = fluid.ParamAttr( - name=name + '_offset', - initializer=fluid.initializer.Uniform(low=-bound, high=bound)) - else: - bias_attr = False - elif init == 'google': - n = filter_size * filter_size * fan_out - param_attr = fluid.ParamAttr( - name=name + "_weights", - initializer=fluid.initializer.NormalInitializer( - loc=0.0, scale=math.sqrt(2.0 / n))) - if use_bias == True: - bias_attr = fluid.ParamAttr( - name=name + "_offset", - initializer=fluid.initializer.Constant(0.0)) - else: - bias_attr = False - - else: - param_attr = fluid.ParamAttr( - name=name + "_weights", - initializer=fluid.initializer.NormalInitializer( - loc=0.0, scale=stddev)) - if use_bias == True: - bias_attr = fluid.ParamAttr( - name=name + "_offset", - initializer=fluid.initializer.Constant(0.0)) - else: - bias_attr = False - return param_attr, bias_attr - - -def cal_padding(img_size, stride, filter_size, dilation=1): - """Calculate padding size.""" - if img_size % stride == 0: - out_size = max(filter_size - stride, 0) - else: - out_size = max(filter_size - (img_size % stride), 0) - return out_size // 2, out_size - out_size // 2 - - -def init_batch_norm_layer(name="batch_norm"): - param_attr = fluid.ParamAttr( - name=name + '_scale', initializer=fluid.initializer.Constant(1.0)) - bias_attr = fluid.ParamAttr( - name=name + '_offset', - initializer=fluid.initializer.Constant(value=0.0)) - return param_attr, bias_attr - - -def init_fc_layer(fout, name='fc'): - n = fout # fan-out - init_range = 1.0 / math.sqrt(n) - - param_attr = fluid.ParamAttr( - name=name + '_weights', - initializer=fluid.initializer.UniformInitializer( - low=-init_range, high=init_range)) - bias_attr = fluid.ParamAttr( - name=name + '_offset', - initializer=fluid.initializer.Constant(value=0.0)) - return param_attr, bias_attr - - -def norm_layer(input, norm_type='batch_norm', name=None): - if norm_type == 'batch_norm': - param_attr = fluid.ParamAttr( - name=name + '_weights', initializer=fluid.initializer.Constant(1.0)) - bias_attr = fluid.ParamAttr( - name=name + '_offset', - initializer=fluid.initializer.Constant(value=0.0)) - return fluid.layers.batch_norm( - input, - param_attr=param_attr, - bias_attr=bias_attr, - moving_mean_name=name + '_mean', - moving_variance_name=name + '_variance') - - elif norm_type == 'instance_norm': - helper = fluid.layer_helper.LayerHelper("instance_norm", **locals()) - dtype = helper.input_dtype() - epsilon = 1e-5 - mean = fluid.layers.reduce_mean(input, dim=[2, 3], keep_dim=True) - var = fluid.layers.reduce_mean( - fluid.layers.square(input - mean), dim=[2, 3], keep_dim=True) - if name is not None: - scale_name = name + "_scale" - offset_name = name + "_offset" - scale_param = fluid.ParamAttr( - name=scale_name, - initializer=fluid.initializer.Constant(1.0), - trainable=True) - offset_param = fluid.ParamAttr( - name=offset_name, - initializer=fluid.initializer.Constant(0.0), - trainable=True) - scale = helper.create_parameter( - attr=scale_param, shape=input.shape[1:2], dtype=dtype) - offset = helper.create_parameter( - attr=offset_param, shape=input.shape[1:2], dtype=dtype) - - tmp = fluid.layers.elementwise_mul(x=(input - mean), y=scale, axis=1) - tmp = tmp / fluid.layers.sqrt(var + epsilon) - tmp = fluid.layers.elementwise_add(tmp, offset, axis=1) - return tmp - else: - raise NotImplementedError("norm tyoe: [%s] is not support" % norm_type) - - -def conv2d(input, - num_filters=64, - filter_size=7, - stride=1, - stddev=0.02, - padding=0, - groups=None, - name="conv2d", - norm=None, - act=None, - relufactor=0.0, - use_bias=False, - padding_type=None, - initial="normal", - use_cudnn=True): - - if padding != 0 and padding_type != None: - warnings.warn( - 'padding value and padding type are set in the same time, and the final padding width and padding height are computed by padding_type' - ) - - param_attr, bias_attr = initial_type( - name=name, - input=input, - op_type='conv', - fan_out=num_filters, - init=initial, - use_bias=use_bias, - filter_size=filter_size, - stddev=stddev) - - def get_padding(filter_size, stride=1, dilation=1): - padding = ((stride - 1) + dilation * (filter_size - 1)) // 2 - return padding - - need_crop = False - if padding_type == "SAME": - top_padding, bottom_padding = cal_padding(input.shape[2], stride, - filter_size) - left_padding, right_padding = cal_padding(input.shape[2], stride, - filter_size) - height_padding = bottom_padding - width_padding = right_padding - if top_padding != bottom_padding or left_padding != right_padding: - height_padding = top_padding + stride - width_padding = left_padding + stride - need_crop = True - padding = [height_padding, width_padding] - elif padding_type == "VALID": - height_padding = 0 - width_padding = 0 - padding = [height_padding, width_padding] - elif padding_type == "DYNAMIC": - padding = get_padding(filter_size, stride) - else: - padding = padding - - conv = fluid.layers.conv2d( - input, - num_filters, - filter_size, - groups=groups, - name=name, - stride=stride, - padding=padding, - use_cudnn=use_cudnn, - param_attr=param_attr, - bias_attr=bias_attr) - - if need_crop: - conv = conv[:, :, 1:, 1:] - - if norm is not None: - conv = norm_layer(input=conv, norm_type=norm, name=name + "_norm") - if act == 'relu': - conv = fluid.layers.relu(conv, name=name + '_relu') - elif act == 'leaky_relu': - conv = fluid.layers.leaky_relu( - conv, alpha=relufactor, name=name + '_leaky_relu') - elif act == 'tanh': - conv = fluid.layers.tanh(conv, name=name + '_tanh') - elif act == 'sigmoid': - conv = fluid.layers.sigmoid(conv, name=name + '_sigmoid') - elif act == 'swish': - conv = fluid.layers.swish(conv, name=name + '_swish') - elif act == None: - conv = conv - else: - raise NotImplementedError("activation: [%s] is not support" % act) - - return conv diff --git a/hub_module/modules/image/classification/efficientnetb2_imagenet/module.py b/hub_module/modules/image/classification/efficientnetb2_imagenet/module.py index 11ccf5e5..6b92e01c 100644 --- a/hub_module/modules/image/classification/efficientnetb2_imagenet/module.py +++ b/hub_module/modules/image/classification/efficientnetb2_imagenet/module.py @@ -1,4 +1,3 @@ -# -*- coding:utf-8 -*- # copyright (c) 2020 PaddlePaddle Authors. All Rights Reserve. # # Licensed under the Apache License, Version 2.0 (the "License"); @@ -13,318 +12,747 @@ # See the License for the specific language governing permissions and # limitations under the License. -from __future__ import absolute_import -from __future__ import division - -import ast -import argparse import os - -import numpy as np -import paddle.fluid as fluid -import paddlehub as hub -from paddle.fluid.core import PaddleTensor, AnalysisConfig, create_paddle_predictor -from paddlehub.module.module import moduleinfo, runnable, serving -from paddlehub.common.paddle_helper import add_vars_prefix - -from efficientnetb2_imagenet.processor import postprocess, base64_to_cv2 -from efficientnetb2_imagenet.data_feed import reader -from efficientnetb2_imagenet.efficientnet import EfficientNetB2 - - -@moduleinfo( - name="efficientnetb2_imagenet", - type="CV/image_classification", - author="paddlepaddle", - author_email="paddle-dev@baidu.com", - summary= - "EfficientNetB2 is a image classfication model, this module is trained with imagenet datasets.", - version="1.1.0") -class EfficientNetB2ImageNet(hub.Module): - def _initialize(self): - self.default_pretrained_model_path = os.path.join( - self.directory, "efficientnetb2_imagenet_infer_model") - label_file = os.path.join(self.directory, "label_list.txt") - with open(label_file, 'r', encoding='utf-8') as file: - self.label_list = file.read().split("\n")[:-1] - self.classification = self.classify - self._set_config() - - def get_expected_image_width(self): - return 224 - - def get_expected_image_height(self): - return 224 - - def get_pretrained_images_mean(self): - im_mean = np.array([0.485, 0.456, 0.406]).reshape(1, 3) - return im_mean - - def get_pretrained_images_std(self): - im_std = np.array([0.229, 0.224, 0.225]).reshape(1, 3) - return im_std - - def _set_config(self): - """ - predictor config setting - """ - cpu_config = AnalysisConfig(self.default_pretrained_model_path) - cpu_config.disable_glog_info() - cpu_config.disable_gpu() - self.cpu_predictor = create_paddle_predictor(cpu_config) - - try: - _places = os.environ["CUDA_VISIBLE_DEVICES"] - int(_places[0]) - use_gpu = True - except: - use_gpu = False - if use_gpu: - gpu_config = AnalysisConfig(self.default_pretrained_model_path) - gpu_config.disable_glog_info() - gpu_config.enable_use_gpu( - memory_pool_init_size_mb=1000, device_id=0) - self.gpu_predictor = create_paddle_predictor(gpu_config) - - def context(self, - trainable=True, - pretrained=True, - override_params=None, - phase='train'): - """context for transfer learning. - - Args: - trainable (bool): Set parameters in program to be trainable. - pretrained (bool) : Whether to load pretrained model. - - Returns: - inputs (dict): key is 'image', corresponding vaule is image tensor. - outputs (dict): key is : - 'classification', corresponding value is the result of classification. - 'feature_map', corresponding value is the result of the layer before the fully connected layer. - context_prog (fluid.Program): program for transfer learning. - """ - if phase in ["dev", "test", "predict", "eval"]: - is_test = True - elif phase in ["train"]: - is_test = False - else: - raise ValueError( - "Phase %s is error, which must be one of train, dev, test, eval and predict." - % phase) - - context_prog = fluid.Program() - startup_prog = fluid.Program() - with fluid.program_guard(context_prog, startup_prog): - with fluid.unique_name.guard(): - image = fluid.layers.data( - name="image", shape=[3, 224, 224], dtype="float32") - efficientnet_b2 = EfficientNetB2( - override_params=override_params) - output, feature_map = efficientnet_b2.net( - input=image, - class_dim=len(self.label_list), - is_test=is_test) - - name_prefix = '@HUB_{}@'.format(self.name) - inputs = {'image': name_prefix + image.name} - outputs = { - 'classification': name_prefix + output.name, - 'feature_map': name_prefix + feature_map.name - } - add_vars_prefix(context_prog, name_prefix) - add_vars_prefix(startup_prog, name_prefix) - global_vars = context_prog.global_block().vars - inputs = { - key: global_vars[value] - for key, value in inputs.items() - } - outputs = { - key: global_vars[value] - for key, value in outputs.items() - } - - place = fluid.CPUPlace() - exe = fluid.Executor(place) - # pretrained - if pretrained: - - def _if_exist(var): - b = os.path.exists( - os.path.join(self.default_pretrained_model_path, - var.name)) - return b - - fluid.io.load_vars( - exe, - self.default_pretrained_model_path, - context_prog, - predicate=_if_exist) - else: - exe.run(startup_prog) - # trainable - for param in context_prog.global_block().iter_parameters(): - param.trainable = trainable - return inputs, outputs, context_prog - - def classify(self, - images=None, - paths=None, - batch_size=1, - use_gpu=False, - top_k=1): +import math +import collections +import re +import copy + +import paddle +from paddle import ParamAttr +import paddle.nn as nn +import paddle.nn.functional as F +from paddle.nn import Conv2d, BatchNorm, Linear, Dropout +from paddle.nn import AdaptiveAvgPool2d, MaxPool2d, AvgPool2d +from paddlehub.module.module import moduleinfo +from paddlehub.module.cv_module import ImageClassifierModule + +GlobalParams = collections.namedtuple('GlobalParams', [ + 'batch_norm_momentum', + 'batch_norm_epsilon', + 'dropout_rate', + 'num_classes', + 'width_coefficient', + 'depth_coefficient', + 'depth_divisor', + 'min_depth', + 'drop_connect_rate', +]) + +BlockArgs = collections.namedtuple( + 'BlockArgs', + ['kernel_size', 'num_repeat', 'input_filters', 'output_filters', 'expand_ratio', 'id_skip', 'stride', 'se_ratio']) + +GlobalParams.__new__.__defaults__ = (None, ) * len(GlobalParams._fields) +BlockArgs.__new__.__defaults__ = (None, ) * len(BlockArgs._fields) + + +def efficientnet_params(model_name: str): + """ Map EfficientNet model name to parameter coefficients. """ + params_dict = { + # Coefficients: width,depth,resolution,dropout + 'efficientnet-b2': (1.1, 1.2, 260, 0.3) + } + return params_dict[model_name] + + +def efficientnet(width_coefficient: float = None, + depth_coefficient: float = None, + dropout_rate: float = 0.2, + drop_connect_rate: float = 0.2): + """ Get block arguments according to parameter and coefficients. """ + blocks_args = [ + 'r1_k3_s11_e1_i32_o16_se0.25', + 'r2_k3_s22_e6_i16_o24_se0.25', + 'r2_k5_s22_e6_i24_o40_se0.25', + 'r3_k3_s22_e6_i40_o80_se0.25', + 'r3_k5_s11_e6_i80_o112_se0.25', + 'r4_k5_s22_e6_i112_o192_se0.25', + 'r1_k3_s11_e6_i192_o320_se0.25', + ] + blocks_args = BlockDecoder.decode(blocks_args) + + global_params = GlobalParams(batch_norm_momentum=0.99, + batch_norm_epsilon=1e-3, + dropout_rate=dropout_rate, + drop_connect_rate=drop_connect_rate, + num_classes=1000, + width_coefficient=width_coefficient, + depth_coefficient=depth_coefficient, + depth_divisor=8, + min_depth=None) + + return blocks_args, global_params + + +def get_model_params(model_name: str, override_params: dict): + """ Get the block args and global params for a given model """ + if model_name.startswith('efficientnet'): + w, d, _, p = efficientnet_params(model_name) + blocks_args, global_params = efficientnet(width_coefficient=w, depth_coefficient=d, dropout_rate=p) + else: + raise NotImplementedError('model name is not pre-defined: %s' % model_name) + if override_params: + global_params = global_params._replace(**override_params) + return blocks_args, global_params + + +def round_filters(filters: int, global_params: dict): + """ Calculate and round number of filters based on depth multiplier. """ + multiplier = global_params.width_coefficient + if not multiplier: + return filters + divisor = global_params.depth_divisor + min_depth = global_params.min_depth + filters *= multiplier + min_depth = min_depth or divisor + new_filters = max(min_depth, int(filters + divisor / 2) // divisor * divisor) + if new_filters < 0.9 * filters: # prevent rounding by more than 10% + new_filters += divisor + return int(new_filters) + + +def round_repeats(repeats: int, global_params: dict): + """ Round number of filters based on depth multiplier. """ + multiplier = global_params.depth_coefficient + if not multiplier: + return repeats + return int(math.ceil(multiplier * repeats)) + + +class BlockDecoder(object): + """ + Block Decoder, straight from the official TensorFlow repository. + """ + @staticmethod + def _decode_block_string(block_string: str): + """ Gets a block through a string notation of arguments. """ + assert isinstance(block_string, str) + + ops = block_string.split('_') + options = {} + for op in ops: + splits = re.split(r'(\d.*)', op) + if len(splits) >= 2: + key, value = splits[:2] + options[key] = value + + # Check stride + cond_1 = ('s' in options and len(options['s']) == 1) + cond_2 = ((len(options['s']) == 2) and (options['s'][0] == options['s'][1])) + assert (cond_1 or cond_2) + + return BlockArgs(kernel_size=int(options['k']), + num_repeat=int(options['r']), + input_filters=int(options['i']), + output_filters=int(options['o']), + expand_ratio=int(options['e']), + id_skip=('noskip' not in block_string), + se_ratio=float(options['se']) if 'se' in options else None, + stride=[int(options['s'][0])]) + + @staticmethod + def _encode_block_string(block): + """Encodes a block to a string.""" + args = [ + 'r%d' % block.num_repeat, + 'k%d' % block.kernel_size, + 's%d%d' % (block.strides[0], block.strides[1]), + 'e%s' % block.expand_ratio, + 'i%d' % block.input_filters, + 'o%d' % block.output_filters + ] + if 0 < block.se_ratio <= 1: + args.append('se%s' % block.se_ratio) + if block.id_skip is False: + args.append('noskip') + return '_'.join(args) + + @staticmethod + def decode(string_list: list): """ - API for image classification. - - Args: - images (list[numpy.ndarray]): data of images, shape of each is [H, W, C], color space must be BGR. - paths (list[str]): The paths of images. - batch_size (int): batch size. - use_gpu (bool): Whether to use gpu. - top_k (int): Return top k results. + Decode a list of string notations to specify blocks in the network. - Returns: - res (list[dict]): The classfication results. + string_list: list of strings, each string is a notation of block + return + list of BlockArgs namedtuples of block args """ - if use_gpu: - try: - _places = os.environ["CUDA_VISIBLE_DEVICES"] - int(_places[0]) - except: - raise RuntimeError( - "Environment Variable CUDA_VISIBLE_DEVICES is not set correctly. If you wanna use gpu, please set CUDA_VISIBLE_DEVICES as cuda_device_id." - ) - - all_data = list() - for yield_data in reader(images, paths): - all_data.append(yield_data) - - total_num = len(all_data) - loop_num = int(np.ceil(total_num / batch_size)) - - res = list() - for iter_id in range(loop_num): - batch_data = list() - handle_id = iter_id * batch_size - for image_id in range(batch_size): - try: - batch_data.append(all_data[handle_id + image_id]) - except: - pass - # feed batch image - batch_image = np.array([data['image'] for data in batch_data]) - batch_image = PaddleTensor(batch_image.copy()) - predictor_output = self.gpu_predictor.run([ - batch_image - ]) if use_gpu else self.cpu_predictor.run([batch_image]) - out = postprocess( - data_out=predictor_output[0].as_ndarray(), - label_list=self.label_list, - top_k=top_k) - res += out - return res - - def save_inference_model(self, - dirname, - model_filename=None, - params_filename=None, - combined=True): - if combined: - model_filename = "__model__" if not model_filename else model_filename - params_filename = "__params__" if not params_filename else params_filename - place = fluid.CPUPlace() - exe = fluid.Executor(place) - - program, feeded_var_names, target_vars = fluid.io.load_inference_model( - dirname=self.default_pretrained_model_path, executor=exe) - - fluid.io.save_inference_model( - dirname=dirname, - main_program=program, - executor=exe, - feeded_var_names=feeded_var_names, - target_vars=target_vars, - model_filename=model_filename, - params_filename=params_filename) - - @serving - def serving_method(self, images, **kwargs): + assert isinstance(string_list, list) + blocks_args = [] + for block_string in string_list: + blocks_args.append(BlockDecoder._decode_block_string(block_string)) + return blocks_args + + @staticmethod + def encode(blocks_args: list): """ - Run as a service. - """ - images_decode = [base64_to_cv2(image) for image in images] - results = self.classify(images=images_decode, **kwargs) - return results + Encodes a list of BlockArgs to a list of strings. - @runnable - def run_cmd(self, argvs): - """ - Run as a command. + :param blocks_args: a list of BlockArgs namedtuples of block args + :return: a list of strings, each string is a notation of block """ - self.parser = argparse.ArgumentParser( - description="Run the {} module.".format(self.name), - prog='hub run {}'.format(self.name), - usage='%(prog)s', - add_help=True) - self.arg_input_group = self.parser.add_argument_group( - title="Input options", description="Input data. Required") - self.arg_config_group = self.parser.add_argument_group( - title="Config options", - description= - "Run configuration for controlling module behavior, not required.") - self.add_module_config_arg() - self.add_module_input_arg() - args = self.parser.parse_args(argvs) - results = self.classify( - paths=[args.input_path], - batch_size=args.batch_size, - use_gpu=args.use_gpu) - return results - - def add_module_config_arg(self): - """ - Add the command config options. - """ - self.arg_config_group.add_argument( - '--use_gpu', - type=ast.literal_eval, - default=False, - help="whether use GPU or not.") - self.arg_config_group.add_argument( - '--batch_size', - type=ast.literal_eval, - default=1, - help="batch size.") - self.arg_config_group.add_argument( - '--top_k', - type=ast.literal_eval, - default=1, - help="Return top k results.") - - def add_module_input_arg(self): - """ - Add the command input options. - """ - self.arg_input_group.add_argument( - '--input_path', type=str, help="path to image.") - - -if __name__ == '__main__': - b2 = EfficientNetB2ImageNet() - b2.context() - import cv2 - test_image = [ - cv2.imread( - '/mnt/zhangxuefei/program-paddle/PaddleHub/hub_module/tests/image_dataset/classification/animals/dog.jpeg' - ) - ] - res = b2.classification(images=test_image) - print(res) - res = b2.classification(paths=[ - '/mnt/zhangxuefei/program-paddle/PaddleHub/hub_module/tests/image_dataset/classification/animals/dog.jpeg' - ]) - print(res) - res = b2.classification(images=test_image) - print(res) - res = b2.classify(images=test_image) - print(res) + block_strings = [] + for block in blocks_args: + block_strings.append(BlockDecoder._encode_block_string(block)) + return block_strings + + +def initial_type(name: str, use_bias: bool = False): + param_attr = ParamAttr(name=name + "_weights") + if use_bias: + bias_attr = ParamAttr(name=name + "_offset") + else: + bias_attr = False + return param_attr, bias_attr + + +def init_batch_norm_layer(name: str = "batch_norm"): + param_attr = ParamAttr(name=name + "_scale") + bias_attr = ParamAttr(name=name + "_offset") + return param_attr, bias_attr + + +def init_fc_layer(name: str = "fc"): + param_attr = ParamAttr(name=name + "_weights") + bias_attr = ParamAttr(name=name + "_offset") + return param_attr, bias_attr + + +def cal_padding(img_size: int, stride: int, filter_size: int, dilation: int = 1): + """Calculate padding size.""" + if img_size % stride == 0: + out_size = max(filter_size - stride, 0) + else: + out_size = max(filter_size - (img_size % stride), 0) + return out_size // 2, out_size - out_size // 2 + + +inp_shape = {"b2": [260, 130, 130, 65, 33, 17, 17, 9]} + + +def _drop_connect(inputs: paddle.Tensor, prob: float, is_test: bool): + """Drop input connection""" + if is_test: + return inputs + keep_prob = 1.0 - prob + inputs_shape = paddle.shape(inputs) + random_tensor = keep_prob + paddle.rand(shape=[inputs_shape[0], 1, 1, 1]) + binary_tensor = paddle.floor(random_tensor) + output = inputs / keep_prob * binary_tensor + return output + + +class Conv2ds(nn.Layer): + """Basic conv layer""" + def __init__(self, + input_channels: int, + output_channels: int, + filter_size: int, + stride: int = 1, + padding: int = 0, + groups: int = None, + name: str = "conv2d", + act: str = None, + use_bias: bool = False, + padding_type: str = None, + model_name: str = None, + cur_stage: str = None): + super(Conv2ds, self).__init__() + assert act in [None, "swish", "sigmoid"] + self.act = act + + param_attr, bias_attr = initial_type(name=name, use_bias=use_bias) + + def get_padding(filter_size, stride=1, dilation=1): + padding = ((stride - 1) + dilation * (filter_size - 1)) // 2 + return padding + + inps = 1 if model_name == None and cur_stage == None else inp_shape[model_name][cur_stage] + self.need_crop = False + if padding_type == "SAME": + top_padding, bottom_padding = cal_padding(inps, stride, filter_size) + left_padding, right_padding = cal_padding(inps, stride, filter_size) + height_padding = bottom_padding + width_padding = right_padding + if top_padding != bottom_padding or left_padding != right_padding: + height_padding = top_padding + stride + width_padding = left_padding + stride + self.need_crop = True + padding = [height_padding, width_padding] + elif padding_type == "VALID": + height_padding = 0 + width_padding = 0 + padding = [height_padding, width_padding] + elif padding_type == "DYNAMIC": + padding = get_padding(filter_size, stride) + else: + padding = padding_type + + groups = 1 if groups is None else groups + self._conv = Conv2d(input_channels, + output_channels, + filter_size, + groups=groups, + stride=stride, + padding=padding, + weight_attr=param_attr, + bias_attr=bias_attr) + + def forward(self, inputs: paddle.Tensor): + x = self._conv(inputs) + if self.act == "swish": + x = F.swish(x) + elif self.act == "sigmoid": + x = F.sigmoid(x) + + if self.need_crop: + x = x[:, :, 1:, 1:] + return x + + +class ConvBNLayer(nn.Layer): + """Basic conv bn layer.""" + def __init__(self, + input_channels: int, + filter_size: int, + output_channels: int, + stride: int = 1, + num_groups: int = 1, + padding_type: str = "SAME", + conv_act: str = None, + bn_act: str = "swish", + use_bn: bool = True, + use_bias: bool = False, + name: str = None, + conv_name: str = None, + bn_name: str = None, + model_name: str = None, + cur_stage: str = None): + super(ConvBNLayer, self).__init__() + + self._conv = Conv2ds(input_channels=input_channels, + output_channels=output_channels, + filter_size=filter_size, + stride=stride, + groups=num_groups, + act=conv_act, + padding_type=padding_type, + name=conv_name, + use_bias=use_bias, + model_name=model_name, + cur_stage=cur_stage) + self.use_bn = use_bn + if use_bn is True: + bn_name = name + bn_name + param_attr, bias_attr = init_batch_norm_layer(bn_name) + + self._bn = BatchNorm(num_channels=output_channels, + act=bn_act, + momentum=0.99, + epsilon=0.001, + moving_mean_name=bn_name + "_mean", + moving_variance_name=bn_name + "_variance", + param_attr=param_attr, + bias_attr=bias_attr) + + def forward(self, inputs: paddle.Tensor): + if self.use_bn: + x = self._conv(inputs) + x = self._bn(x) + return x + else: + return self._conv(inputs) + + +class ExpandConvNorm(nn.Layer): + """Expand conv norm layer.""" + def __init__(self, + input_channels: int, + block_args: dict, + padding_type: str, + name: str = None, + model_name: str = None, + cur_stage: str = None): + super(ExpandConvNorm, self).__init__() + + self.oup = block_args.input_filters * block_args.expand_ratio + self.expand_ratio = block_args.expand_ratio + + if self.expand_ratio != 1: + self._conv = ConvBNLayer(input_channels, + 1, + self.oup, + bn_act=None, + padding_type=padding_type, + name=name, + conv_name=name + "_expand_conv", + bn_name="_bn0", + model_name=model_name, + cur_stage=cur_stage) + + def forward(self, inputs: paddle.Tensor): + if self.expand_ratio != 1: + return self._conv(inputs) + else: + return inputs + + +class DepthwiseConvNorm(nn.Layer): + """Depthwise conv norm layer.""" + def __init__(self, + input_channels: int, + block_args: dict, + padding_type: str, + name: str = None, + model_name: str = None, + cur_stage: str = None): + super(DepthwiseConvNorm, self).__init__() + + self.k = block_args.kernel_size + self.s = block_args.stride + if isinstance(self.s, list) or isinstance(self.s, tuple): + self.s = self.s[0] + oup = block_args.input_filters * block_args.expand_ratio + + self._conv = ConvBNLayer(input_channels, + self.k, + oup, + self.s, + num_groups=input_channels, + bn_act=None, + padding_type=padding_type, + name=name, + conv_name=name + "_depthwise_conv", + bn_name="_bn1", + model_name=model_name, + cur_stage=cur_stage) + + def forward(self, inputs: paddle.Tensor): + return self._conv(inputs) + + +class ProjectConvNorm(nn.Layer): + """Projection conv bn layer.""" + def __init__(self, + input_channels: int, + block_args: dict, + padding_type: str, + name: str = None, + model_name: str = None, + cur_stage: str = None): + super(ProjectConvNorm, self).__init__() + + final_oup = block_args.output_filters + + self._conv = ConvBNLayer(input_channels, + 1, + final_oup, + bn_act=None, + padding_type=padding_type, + name=name, + conv_name=name + "_project_conv", + bn_name="_bn2", + model_name=model_name, + cur_stage=cur_stage) + + def forward(self, inputs: paddle.Tensor): + return self._conv(inputs) + + +class SEBlock(nn.Layer): + """Basic Squeeze-and-Excitation block for Efficientnet.""" + def __init__(self, + input_channels: int, + num_squeezed_channels: int, + oup: int, + padding_type: str, + name: str = None, + model_name: str = None, + cur_stage: str = None): + super(SEBlock, self).__init__() + + self._pool = AdaptiveAvgPool2d(1) + self._conv1 = Conv2ds(input_channels, + num_squeezed_channels, + 1, + use_bias=True, + padding_type=padding_type, + act="swish", + name=name + "_se_reduce") + + self._conv2 = Conv2ds(num_squeezed_channels, + oup, + 1, + act="sigmoid", + use_bias=True, + padding_type=padding_type, + name=name + "_se_expand") + + def forward(self, inputs: paddle.Tensor): + x = self._pool(inputs) + x = self._conv1(x) + x = self._conv2(x) + return paddle.multiply(inputs, x) + + +class MbConvBlock(nn.Layer): + """Mobile inverted bottleneck convolution for Efficientnet.""" + def __init__(self, + input_channels: int, + block_args: dict, + padding_type: str, + use_se: bool, + name: str = None, + drop_connect_rate: float = None, + is_test: bool = False, + model_name: str = None, + cur_stage: str = None): + super(MbConvBlock, self).__init__() + + oup = block_args.input_filters * block_args.expand_ratio + self.block_args = block_args + self.has_se = use_se and (block_args.se_ratio is not None) and (0 < block_args.se_ratio <= 1) + self.id_skip = block_args.id_skip + self.expand_ratio = block_args.expand_ratio + self.drop_connect_rate = drop_connect_rate + self.is_test = is_test + + if self.expand_ratio != 1: + self._ecn = ExpandConvNorm(input_channels, + block_args, + padding_type=padding_type, + name=name, + model_name=model_name, + cur_stage=cur_stage) + + self._dcn = DepthwiseConvNorm(input_channels * block_args.expand_ratio, + block_args, + padding_type=padding_type, + name=name, + model_name=model_name, + cur_stage=cur_stage) + + if self.has_se: + num_squeezed_channels = max(1, int(block_args.input_filters * block_args.se_ratio)) + self._se = SEBlock(input_channels * block_args.expand_ratio, + num_squeezed_channels, + oup, + padding_type=padding_type, + name=name, + model_name=model_name, + cur_stage=cur_stage) + + self._pcn = ProjectConvNorm(input_channels * block_args.expand_ratio, + block_args, + padding_type=padding_type, + name=name, + model_name=model_name, + cur_stage=cur_stage) + + def forward(self, inputs: paddle.Tensor): + x = inputs + if self.expand_ratio != 1: + x = self._ecn(x) + x = F.swish(x) + x = self._dcn(x) + x = F.swish(x) + if self.has_se: + x = self._se(x) + x = self._pcn(x) + if self.id_skip and \ + self.block_args.stride == 1 and \ + self.block_args.input_filters == self.block_args.output_filters: + if self.drop_connect_rate: + x = _drop_connect(x, self.drop_connect_rate, self.is_test) + x = paddle.elementwise_add(x, inputs) + return x + + +class ConvStemNorm(nn.Layer): + """Basic conv stem norm block for extracting features.""" + def __init__(self, + input_channels: int, + padding_type: str, + _global_params: dict, + name: str = None, + model_name: str = None, + cur_stage: str = None): + super(ConvStemNorm, self).__init__() + + output_channels = round_filters(32, _global_params) + self._conv = ConvBNLayer(input_channels, + filter_size=3, + output_channels=output_channels, + stride=2, + bn_act=None, + padding_type=padding_type, + name="", + conv_name="_conv_stem", + bn_name="_bn0", + model_name=model_name, + cur_stage=cur_stage) + + def forward(self, inputs: paddle.Tensor): + return self._conv(inputs) + + +class ExtractFeatures(nn.Layer): + """Extract features.""" + def __init__(self, + input_channels: int, + _block_args: dict, + _global_params: dict, + padding_type: str, + use_se: bool, + is_test: bool, + model_name: str = None): + super(ExtractFeatures, self).__init__() + + self._global_params = _global_params + + self._conv_stem = ConvStemNorm(input_channels, + padding_type=padding_type, + _global_params=_global_params, + model_name=model_name, + cur_stage=0) + + self.block_args_copy = copy.deepcopy(_block_args) + idx = 0 + block_size = 0 + for block_arg in self.block_args_copy: + block_arg = block_arg._replace(input_filters=round_filters(block_arg.input_filters, _global_params), + output_filters=round_filters(block_arg.output_filters, _global_params), + num_repeat=round_repeats(block_arg.num_repeat, _global_params)) + block_size += 1 + for _ in range(block_arg.num_repeat - 1): + block_size += 1 + + self.conv_seq = [] + cur_stage = 1 + for block_args in _block_args: + block_args = block_args._replace(input_filters=round_filters(block_args.input_filters, _global_params), + output_filters=round_filters(block_args.output_filters, _global_params), + num_repeat=round_repeats(block_args.num_repeat, _global_params)) + + drop_connect_rate = self._global_params.drop_connect_rate if not is_test else 0 + if drop_connect_rate: + drop_connect_rate *= float(idx) / block_size + + _mc_block = self.add_sublayer( + "_blocks." + str(idx) + ".", + MbConvBlock(block_args.input_filters, + block_args=block_args, + padding_type=padding_type, + use_se=use_se, + name="_blocks." + str(idx) + ".", + drop_connect_rate=drop_connect_rate, + model_name=model_name, + cur_stage=cur_stage)) + self.conv_seq.append(_mc_block) + idx += 1 + if block_args.num_repeat > 1: + block_args = block_args._replace(input_filters=block_args.output_filters, stride=1) + for _ in range(block_args.num_repeat - 1): + drop_connect_rate = self._global_params.drop_connect_rate if not is_test else 0 + if drop_connect_rate: + drop_connect_rate *= float(idx) / block_size + _mc_block = self.add_sublayer( + "block." + str(idx) + ".", + MbConvBlock(block_args.input_filters, + block_args, + padding_type=padding_type, + use_se=use_se, + name="_blocks." + str(idx) + ".", + drop_connect_rate=drop_connect_rate, + model_name=model_name, + cur_stage=cur_stage)) + self.conv_seq.append(_mc_block) + idx += 1 + cur_stage += 1 + + def forward(self, inputs: paddle.Tensor): + x = self._conv_stem(inputs) + x = F.swish(x) + for _mc_block in self.conv_seq: + x = _mc_block(x) + return x + + +@moduleinfo(name="efficientnetb2_imagenet", + type="cv/classification", + author="paddlepaddle", + author_email="", + summary="efficientnetb2_imagenet is a classification model, " + "this module is trained with Imagenet dataset.", + version="1.1.0", + meta=ImageClassifierModule) +class EfficientNet_B2(nn.Layer): + def __init__(self, + is_test: bool = False, + padding_type: str = "SAME", + override_params: dict = None, + use_se: bool = True, + class_dim: int = 1000, + load_checkpoint: str = None): + super(EfficientNet_B2, self).__init__() + + model_name = 'efficientnet-b2' + self.name = "b2" + self._block_args, self._global_params = get_model_params(model_name, override_params) + self.padding_type = padding_type + self.use_se = use_se + self.is_test = is_test + + self._ef = ExtractFeatures(3, + self._block_args, + self._global_params, + self.padding_type, + self.use_se, + self.is_test, + model_name=self.name) + + output_channels = round_filters(1280, self._global_params) + oup = 352 + + self._conv = ConvBNLayer(oup, + 1, + output_channels, + bn_act="swish", + padding_type=self.padding_type, + name="", + conv_name="_conv_head", + bn_name="_bn1", + model_name=self.name, + cur_stage=7) + self._pool = AdaptiveAvgPool2d(1) + + if self._global_params.dropout_rate: + self._drop = Dropout(p=self._global_params.dropout_rate, mode="upscale_in_train") + + param_attr, bias_attr = init_fc_layer("_fc") + self._fc = Linear(output_channels, class_dim, weight_attr=param_attr, bias_attr=bias_attr) + + if load_checkpoint is not None: + model_dict = paddle.load(load_checkpoint)[0] + self.set_dict(model_dict) + print("load custom checkpoint success") + + else: + checkpoint = os.path.join(self.directory, 'efficientnet_b2_imagenet.pdparams') + if not os.path.exists(checkpoint): + os.system( + 'wget https://paddlehub.bj.bcebos.com/dygraph/image_classification/efficientnet_b2_imagenet.pdparams -O ' + + checkpoint) + model_dict = paddle.load(checkpoint)[0] + self.set_dict(model_dict) + print("load pretrained checkpoint success") + + def forward(self, inputs: paddle.Tensor): + x = self._ef(inputs) + x = self._conv(x) + x = self._pool(x) + if self._global_params.dropout_rate: + x = self._drop(x) + x = paddle.squeeze(x, axis=[2, 3]) + x = self._fc(x) + return x diff --git a/hub_module/modules/image/classification/efficientnetb2_imagenet/processor.py b/hub_module/modules/image/classification/efficientnetb2_imagenet/processor.py deleted file mode 100644 index 12ccc9b2..00000000 --- a/hub_module/modules/image/classification/efficientnetb2_imagenet/processor.py +++ /dev/null @@ -1,69 +0,0 @@ -# -*- coding:utf-8 -*- -# copyright (c) 2020 PaddlePaddle Authors. All Rights Reserve. -# -# Licensed under the Apache License, Version 2.0 (the "License"); -# you may not use this file except in compliance with the License. -# You may obtain a copy of the License at -# -# http://www.apache.org/licenses/LICENSE-2.0 -# -# Unless required by applicable law or agreed to in writing, software -# distributed under the License is distributed on an "AS IS" BASIS, -# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. -# See the License for the specific language governing permissions and -# limitations under the License. - -from __future__ import absolute_import -from __future__ import division -from __future__ import print_function - -import base64 -import cv2 -import os - -import numpy as np - - -def base64_to_cv2(b64str): - data = base64.b64decode(b64str.encode('utf8')) - data = np.fromstring(data, np.uint8) - data = cv2.imdecode(data, cv2.IMREAD_COLOR) - return data - - -def softmax(x): - orig_shape = x.shape - if len(x.shape) > 1: - tmp = np.max(x, axis=1) - x -= tmp.reshape((x.shape[0], 1)) - x = np.exp(x) - tmp = np.sum(x, axis=1) - x /= tmp.reshape((x.shape[0], 1)) - else: - tmp = np.max(x) - x -= tmp - x = np.exp(x) - tmp = np.sum(x) - x /= tmp - return x - - -def postprocess(data_out, label_list, top_k): - """ - Postprocess output of network, one image at a time. - - Args: - data_out (numpy.ndarray): output data of network. - label_list (list): list of label. - top_k (int): Return top k results. - """ - output = [] - for result in data_out: - result_i = softmax(result) - output_i = {} - indexs = np.argsort(result_i)[::-1][0:top_k] - for index in indexs: - label = label_list[index].split(',')[0] - output_i[label] = float(result_i[index]) - output.append(output_i) - return output diff --git a/hub_module/modules/image/classification/efficientnetb3_imagenet/README.md b/hub_module/modules/image/classification/efficientnetb3_imagenet/README.md deleted file mode 100644 index b6b94bbb..00000000 --- a/hub_module/modules/image/classification/efficientnetb3_imagenet/README.md +++ /dev/null @@ -1,149 +0,0 @@ -## 命令行预测 - -``` -hub run efficientnetb3_imagenet --input_path "/PATH/TO/IMAGE" -``` - -## API - -```python -def get_expected_image_width() -``` - -返回预处理的图片宽度，也就是224。 - -```python -def get_expected_image_height() -``` - -返回预处理的图片高度，也就是224。 - -```python -def get_pretrained_images_mean() -``` - -返回预处理的图片均值，也就是 \[0.485, 0.456, 0.406\]。 - -```python -def get_pretrained_images_std() -``` - -返回预处理的图片标准差，也就是 \[0.229, 0.224, 0.225\]。 - - -```python -def context(trainable=True, pretrained=True) -``` - -**参数** - -* trainable (bool): 计算图的参数是否为可训练的； -* pretrained (bool): 是否加载默认的预训练模型。 - -**返回** - -* inputs (dict): 计算图的输入，key 为 'image', value 为图片的张量； -* outputs (dict): 计算图的输出，key 为 'classification' 和 'feature_map'，其相应的值为： - * classification (paddle.fluid.framework.Variable): 分类结果，也就是全连接层的输出； - * feature\_map (paddle.fluid.framework.Variable): 特征匹配，全连接层前面的那个张量。 -* context\_prog(fluid.Program): 计算图，用于迁移学习。 - -```python -def classify(images=None, - paths=None, - batch_size=1, - use_gpu=False, - top_k=1): -``` - -**参数** - -* images (list\[numpy.ndarray\]): 图片数据，每一个图片数据的shape 均为 \[H, W, C\]，颜色空间为 BGR； -* paths (list\[str\]): 图片的路径； -* batch\_size (int): batch 的大小； -* use\_gpu (bool): 是否使用 GPU 来预测； -* top\_k (int): 返回预测结果的前 k 个。 - -**返回** - -res (list\[dict\]): 分类结果，列表的每一个元素均为字典，其中 key 为识别动物的类别，value为置信度。 - -```python -def save_inference_model(dirname, - model_filename=None, - params_filename=None, - combined=True) -``` - -将模型保存到指定路径。 - -**参数** - -* dirname: 存在模型的目录名称 -* model\_filename: 模型文件名称，默认为\_\_model\_\_ -* params\_filename: 参数文件名称，默认为\_\_params\_\_(仅当`combined`为True时生效) -* combined: 是否将参数保存到统一的一个文件中 - -## 代码示例 - -```python -import paddlehub as hub -import cv2 - -classifier = hub.Module(name="efficientnetb3_small_imagenet") - -result = classifier.classify(images=[cv2.imread('/PATH/TO/IMAGE')]) -# or -# result = classifier.classify(paths=['/PATH/TO/IMAGE']) -``` - -## 服务部署 - -PaddleHub Serving可以部署一个在线图像识别服务。 - -## 第一步：启动PaddleHub Serving - -运行启动命令： -```shell -$ hub serving start -m efficientnetb3_small_imagenet -``` - -这样就完成了一个在线图像识别服务化API的部署，默认端口号为8866。 - -**NOTE:** 如使用GPU预测，则需要在启动服务之前，请设置CUDA\_VISIBLE\_DEVICES环境变量，否则不用设置。 - -## 第二步：发送预测请求 - -配置好服务端，以下数行代码即可实现发送预测请求，获取预测结果 - -```python -import requests -import json -import cv2 -import base64 - - -def cv2_to_base64(image): - data = cv2.imencode('.jpg', image)[1] - return base64.b64encode(data.tostring()).decode('utf8') - - -# 发送HTTP请求 -data = {'images':[cv2_to_base64(cv2.imread("/PATH/TO/IMAGE"))]} -headers = {"Content-type": "application/json"} -url = "http://127.0.0.1:8866/predict/efficientnetb3_small_imagenet" -r = requests.post(url=url, headers=headers, data=json.dumps(data)) - -# 打印预测结果 -print(r.json()["results"]) -``` - -### 查看代码 - -https://github.com/PaddlePaddle/PaddleClas - -### 依赖 - -paddlepaddle >= 1.6.2 - -paddlehub >= 1.6.0 diff --git a/hub_module/modules/image/classification/efficientnetb3_imagenet/__init__.py b/hub_module/modules/image/classification/efficientnetb3_imagenet/__init__.py deleted file mode 100644 index 19a990da..00000000 --- a/hub_module/modules/image/classification/efficientnetb3_imagenet/__init__.py +++ /dev/null @@ -1 +0,0 @@ -# -*- coding:utf-8 -*- diff --git a/hub_module/modules/image/classification/efficientnetb3_imagenet/data_feed.py b/hub_module/modules/image/classification/efficientnetb3_imagenet/data_feed.py deleted file mode 100644 index 11181458..00000000 --- a/hub_module/modules/image/classification/efficientnetb3_imagenet/data_feed.py +++ /dev/null @@ -1,98 +0,0 @@ -# -*- coding:utf-8 -*- -# copyright (c) 2020 PaddlePaddle Authors. All Rights Reserve. -# -# Licensed under the Apache License, Version 2.0 (the "License"); -# you may not use this file except in compliance with the License. -# You may obtain a copy of the License at -# -# http://www.apache.org/licenses/LICENSE-2.0 -# -# Unless required by applicable law or agreed to in writing, software -# distributed under the License is distributed on an "AS IS" BASIS, -# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. -# See the License for the specific language governing permissions and -# limitations under the License. - -import os -import time -from collections import OrderedDict - -import cv2 -import numpy as np -from PIL import Image - -__all__ = ['reader'] - -DATA_DIM = 224 -img_mean = np.array([0.485, 0.456, 0.406]).reshape((3, 1, 1)) -img_std = np.array([0.229, 0.224, 0.225]).reshape((3, 1, 1)) - - -def resize_short(img, target_size): - percent = float(target_size) / min(img.size[0], img.size[1]) - resized_width = int(round(img.size[0] * percent)) - resized_height = int(round(img.size[1] * percent)) - img = img.resize((resized_width, resized_height), Image.LANCZOS) - return img - - -def crop_image(img, target_size, center): - width, height = img.size - size = target_size - if center == True: - w_start = (width - size) / 2 - h_start = (height - size) / 2 - else: - w_start = np.random.randint(0, width - size + 1) - h_start = np.random.randint(0, height - size + 1) - w_end = w_start + size - h_end = h_start + size - img = img.crop((w_start, h_start, w_end, h_end)) - return img - - -def process_image(img): - img = resize_short(img, target_size=256) - img = crop_image(img, target_size=DATA_DIM, center=True) - if img.mode != 'RGB': - img = img.convert('RGB') - img = np.array(img).astype('float32').transpose((2, 0, 1)) / 255 - img -= img_mean - img /= img_std - return img - - -def reader(images=None, paths=None): - """ - Preprocess to yield image. - - Args: - images (list[numpy.ndarray]): images data, shape of each is [H, W, C]. - paths (list[str]): paths to images. - - Yield: - each (collections.OrderedDict): info of original image, preprocessed image. - """ - component = list() - if paths: - for im_path in paths: - each = OrderedDict() - assert os.path.isfile( - im_path), "The {} isn't a valid file path.".format(im_path) - each['org_im_path'] = im_path - each['org_im'] = Image.open(im_path) - each['org_im_width'], each['org_im_height'] = each['org_im'].size - component.append(each) - if images is not None: - assert type(images), "images is a list." - for im in images: - each = OrderedDict() - each['org_im'] = Image.fromarray(im[:, :, ::-1]) - each['org_im_path'] = 'ndarray_time={}'.format( - round(time.time(), 6) * 1e6) - each['org_im_width'], each['org_im_height'] = each['org_im'].size - component.append(each) - - for element in component: - element['image'] = process_image(element['org_im']) - yield element diff --git a/hub_module/modules/image/classification/efficientnetb3_imagenet/efficientnet.py b/hub_module/modules/image/classification/efficientnetb3_imagenet/efficientnet.py deleted file mode 100644 index 18ddf3c4..00000000 --- a/hub_module/modules/image/classification/efficientnetb3_imagenet/efficientnet.py +++ /dev/null @@ -1,623 +0,0 @@ -# -*- coding:utf-8 -*- -# copyright (c) 2020 PaddlePaddle Authors. All Rights Reserve. -# -# Licensed under the Apache License, Version 2.0 (the "License"); -# you may not use this file except in compliance with the License. -# You may obtain a copy of the License at -# -# http://www.apache.org/licenses/LICENSE-2.0 -# -# Unless required by applicable law or agreed to in writing, software -# distributed under the License is distributed on an "AS IS" BASIS, -# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. -# See the License for the specific language governing permissions and -# limitations under the License. - -from __future__ import absolute_import -from __future__ import division -from __future__ import print_function - -import collections -import re -import math -import copy - -import paddle.fluid as fluid -from efficientnetb3_imagenet.layers import conv2d, init_batch_norm_layer, init_fc_layer - -__all__ = [ - 'EfficientNet', 'EfficientNetB0_small', 'EfficientNetB0', 'EfficientNetB1', - 'EfficientNetB2', 'EfficientNetB3', 'EfficientNetB4', 'EfficientNetB5', - 'EfficientNetB6', 'EfficientNetB7' -] - -GlobalParams = collections.namedtuple('GlobalParams', [ - 'batch_norm_momentum', - 'batch_norm_epsilon', - 'dropout_rate', - 'num_classes', - 'width_coefficient', - 'depth_coefficient', - 'depth_divisor', - 'min_depth', - 'drop_connect_rate', -]) - -BlockArgs = collections.namedtuple('BlockArgs', [ - 'kernel_size', 'num_repeat', 'input_filters', 'output_filters', - 'expand_ratio', 'id_skip', 'stride', 'se_ratio' -]) - -GlobalParams.__new__.__defaults__ = (None, ) * len(GlobalParams._fields) -BlockArgs.__new__.__defaults__ = (None, ) * len(BlockArgs._fields) - - -def efficientnet_params(model_name): - """ Map EfficientNet model name to parameter coefficients. """ - params_dict = { - # Coefficients: width,depth,resolution,dropout - 'efficientnet-b0': (1.0, 1.0, 224, 0.2), - 'efficientnet-b1': (1.0, 1.1, 240, 0.2), - 'efficientnet-b2': (1.1, 1.2, 260, 0.3), - 'efficientnet-b3': (1.2, 1.4, 300, 0.3), - 'efficientnet-b4': (1.4, 1.8, 380, 0.4), - 'efficientnet-b5': (1.6, 2.2, 456, 0.4), - 'efficientnet-b6': (1.8, 2.6, 528, 0.5), - 'efficientnet-b7': (2.0, 3.1, 600, 0.5), - } - return params_dict[model_name] - - -def efficientnet(width_coefficient=None, - depth_coefficient=None, - dropout_rate=0.2, - drop_connect_rate=0.2): - """ Get block arguments according to parameter and coefficients. """ - blocks_args = [ - 'r1_k3_s11_e1_i32_o16_se0.25', - 'r2_k3_s22_e6_i16_o24_se0.25', - 'r2_k5_s22_e6_i24_o40_se0.25', - 'r3_k3_s22_e6_i40_o80_se0.25', - 'r3_k5_s11_e6_i80_o112_se0.25', - 'r4_k5_s22_e6_i112_o192_se0.25', - 'r1_k3_s11_e6_i192_o320_se0.25', - ] - blocks_args = BlockDecoder.decode(blocks_args) - - global_params = GlobalParams( - batch_norm_momentum=0.99, - batch_norm_epsilon=1e-3, - dropout_rate=dropout_rate, - drop_connect_rate=drop_connect_rate, - num_classes=1000, - width_coefficient=width_coefficient, - depth_coefficient=depth_coefficient, - depth_divisor=8, - min_depth=None) - - return blocks_args, global_params - - -def get_model_params(model_name, override_params): - """ Get the block args and global params for a given model """ - if model_name.startswith('efficientnet'): - w, d, _, p = efficientnet_params(model_name) - blocks_args, global_params = efficientnet( - width_coefficient=w, depth_coefficient=d, dropout_rate=p) - else: - raise NotImplementedError( - 'model name is not pre-defined: %s' % model_name) - if override_params: - global_params = global_params._replace(**override_params) - return blocks_args, global_params - - -def round_filters(filters, global_params): - """ Calculate and round number of filters based on depth multiplier. """ - multiplier = global_params.width_coefficient - if not multiplier: - return filters - divisor = global_params.depth_divisor - min_depth = global_params.min_depth - filters *= multiplier - min_depth = min_depth or divisor - new_filters = max(min_depth, - int(filters + divisor / 2) // divisor * divisor) - if new_filters < 0.9 * filters: # prevent rounding by more than 10% - new_filters += divisor - return int(new_filters) - - -def round_repeats(repeats, global_params): - """ Round number of filters based on depth multiplier. """ - multiplier = global_params.depth_coefficient - if not multiplier: - return repeats - return int(math.ceil(multiplier * repeats)) - - -class EfficientNet(): - def __init__(self, - name='b0', - padding_type='SAME', - override_params=None, - is_test=False, - use_se=True): - valid_names = ['b' + str(i) for i in range(8)] - assert name in valid_names, 'efficient name should be in b0~b7' - model_name = 'efficientnet-' + name - self._blocks_args, self._global_params = get_model_params( - model_name, override_params) - self._bn_mom = self._global_params.batch_norm_momentum - self._bn_eps = self._global_params.batch_norm_epsilon - self.padding_type = padding_type - self.use_se = use_se - - def net(self, input, class_dim=1000, is_test=False): - - conv = self.extract_features(input, is_test=is_test) - - out_channels = round_filters(1280, self._global_params) - conv = self.conv_bn_layer( - conv, - num_filters=out_channels, - filter_size=1, - bn_act='swish', - bn_mom=self._bn_mom, - bn_eps=self._bn_eps, - padding_type=self.padding_type, - name='', - conv_name='_conv_head', - bn_name='_bn1') - - pool = fluid.layers.pool2d( - input=conv, pool_type='avg', global_pooling=True, use_cudnn=False) - - if not is_test and self._global_params.dropout_rate: - pool = fluid.layers.dropout( - pool, - self._global_params.dropout_rate, - dropout_implementation='upscale_in_train') - - param_attr, bias_attr = init_fc_layer(class_dim, '_fc') - out = fluid.layers.fc( - pool, - class_dim, - name='_fc', - param_attr=param_attr, - bias_attr=bias_attr) - return out, pool - - def _drop_connect(self, inputs, prob, is_test): - if is_test: - return inputs - keep_prob = 1.0 - prob - random_tensor = keep_prob + fluid.layers.uniform_random_batch_size_like( - inputs, [-1, 1, 1, 1], min=0., max=1.) - binary_tensor = fluid.layers.floor(random_tensor) - output = inputs / keep_prob * binary_tensor - return output - - def _expand_conv_norm(self, inputs, block_args, is_test, name=None): - # Expansion phase - oup = block_args.input_filters * block_args.expand_ratio # number of output channels - - if block_args.expand_ratio != 1: - conv = self.conv_bn_layer( - inputs, - num_filters=oup, - filter_size=1, - bn_act=None, - bn_mom=self._bn_mom, - bn_eps=self._bn_eps, - padding_type=self.padding_type, - name=name, - conv_name=name + '_expand_conv', - bn_name='_bn0') - - return conv - - def _depthwise_conv_norm(self, inputs, block_args, is_test, name=None): - k = block_args.kernel_size - s = block_args.stride - if isinstance(s, list) or isinstance(s, tuple): - s = s[0] - oup = block_args.input_filters * block_args.expand_ratio # number of output channels - - conv = self.conv_bn_layer( - inputs, - num_filters=oup, - filter_size=k, - stride=s, - num_groups=oup, - bn_act=None, - padding_type=self.padding_type, - bn_mom=self._bn_mom, - bn_eps=self._bn_eps, - name=name, - use_cudnn=False, - conv_name=name + '_depthwise_conv', - bn_name='_bn1') - - return conv - - def _project_conv_norm(self, inputs, block_args, is_test, name=None): - final_oup = block_args.output_filters - conv = self.conv_bn_layer( - inputs, - num_filters=final_oup, - filter_size=1, - bn_act=None, - padding_type=self.padding_type, - bn_mom=self._bn_mom, - bn_eps=self._bn_eps, - name=name, - conv_name=name + '_project_conv', - bn_name='_bn2') - return conv - - def conv_bn_layer(self, - input, - filter_size, - num_filters, - stride=1, - num_groups=1, - padding_type="SAME", - conv_act=None, - bn_act='swish', - use_cudnn=True, - use_bn=True, - bn_mom=0.9, - bn_eps=1e-05, - use_bias=False, - name=None, - conv_name=None, - bn_name=None): - conv = conv2d( - input=input, - num_filters=num_filters, - filter_size=filter_size, - stride=stride, - groups=num_groups, - act=conv_act, - padding_type=padding_type, - use_cudnn=use_cudnn, - name=conv_name, - use_bias=use_bias) - - if use_bn == False: - return conv - else: - bn_name = name + bn_name - param_attr, bias_attr = init_batch_norm_layer(bn_name) - return fluid.layers.batch_norm( - input=conv, - act=bn_act, - momentum=bn_mom, - epsilon=bn_eps, - name=bn_name, - moving_mean_name=bn_name + '_mean', - moving_variance_name=bn_name + '_variance', - param_attr=param_attr, - bias_attr=bias_attr) - - def _conv_stem_norm(self, inputs, is_test): - out_channels = round_filters(32, self._global_params) - bn = self.conv_bn_layer( - inputs, - num_filters=out_channels, - filter_size=3, - stride=2, - bn_act=None, - bn_mom=self._bn_mom, - padding_type=self.padding_type, - bn_eps=self._bn_eps, - name='', - conv_name='_conv_stem', - bn_name='_bn0') - - return bn - - def mb_conv_block(self, - inputs, - block_args, - is_test=False, - drop_connect_rate=None, - name=None): - # Expansion and Depthwise Convolution - oup = block_args.input_filters * block_args.expand_ratio # number of output channels - has_se = self.use_se and (block_args.se_ratio is - not None) and (0 < block_args.se_ratio <= 1) - id_skip = block_args.id_skip # skip connection and drop connect - conv = inputs - if block_args.expand_ratio != 1: - conv = fluid.layers.swish( - self._expand_conv_norm(conv, block_args, is_test, name)) - - conv = fluid.layers.swish( - self._depthwise_conv_norm(conv, block_args, is_test, name)) - - # Squeeze and Excitation - if has_se: - num_squeezed_channels = max( - 1, int(block_args.input_filters * block_args.se_ratio)) - conv = self.se_block(conv, num_squeezed_channels, oup, name) - - conv = self._project_conv_norm(conv, block_args, is_test, name) - - # Skip connection and drop connect - input_filters, output_filters = block_args.input_filters, block_args.output_filters - if id_skip and block_args.stride == 1 and input_filters == output_filters: - if drop_connect_rate: - conv = self._drop_connect(conv, drop_connect_rate, is_test) - conv = fluid.layers.elementwise_add(conv, inputs) - - return conv - - def se_block(self, inputs, num_squeezed_channels, oup, name): - x_squeezed = fluid.layers.pool2d( - input=inputs, pool_type='avg', global_pooling=True, use_cudnn=False) - x_squeezed = conv2d( - x_squeezed, - num_filters=num_squeezed_channels, - filter_size=1, - use_bias=True, - padding_type=self.padding_type, - act='swish', - name=name + '_se_reduce') - x_squeezed = conv2d( - x_squeezed, - num_filters=oup, - filter_size=1, - use_bias=True, - padding_type=self.padding_type, - name=name + '_se_expand') - se_out = inputs * fluid.layers.sigmoid(x_squeezed) - return se_out - - def extract_features(self, inputs, is_test): - """ Returns output of the final convolution layer """ - - conv = fluid.layers.swish(self._conv_stem_norm(inputs, is_test=is_test)) - - block_args_copy = copy.deepcopy(self._blocks_args) - idx = 0 - block_size = 0 - for block_arg in block_args_copy: - block_arg = block_arg._replace( - input_filters=round_filters(block_arg.input_filters, - self._global_params), - output_filters=round_filters(block_arg.output_filters, - self._global_params), - num_repeat=round_repeats(block_arg.num_repeat, - self._global_params)) - block_size += 1 - for _ in range(block_arg.num_repeat - 1): - block_size += 1 - - for block_args in self._blocks_args: - - # Update block input and output filters based on depth multiplier. - block_args = block_args._replace( - input_filters=round_filters(block_args.input_filters, - self._global_params), - output_filters=round_filters(block_args.output_filters, - self._global_params), - num_repeat=round_repeats(block_args.num_repeat, - self._global_params)) - - # The first block needs to take care of stride and filter size increase. - drop_connect_rate = self._global_params.drop_connect_rate - if drop_connect_rate: - drop_connect_rate *= float(idx) / block_size - conv = self.mb_conv_block(conv, block_args, is_test, - drop_connect_rate, - '_blocks.' + str(idx) + '.') - - idx += 1 - if block_args.num_repeat > 1: - block_args = block_args._replace( - input_filters=block_args.output_filters, stride=1) - for _ in range(block_args.num_repeat - 1): - drop_connect_rate = self._global_params.drop_connect_rate - if drop_connect_rate: - drop_connect_rate *= float(idx) / block_size - conv = self.mb_conv_block(conv, block_args, is_test, - drop_connect_rate, - '_blocks.' + str(idx) + '.') - idx += 1 - - return conv - - def shortcut(self, input, data_residual): - return fluid.layers.elementwise_add(input, data_residual) - - -class BlockDecoder(object): - """ Block Decoder for readability, straight from the official TensorFlow repository """ - - @staticmethod - def _decode_block_string(block_string): - """ Gets a block through a string notation of arguments. """ - assert isinstance(block_string, str) - - ops = block_string.split('_') - options = {} - for op in ops: - splits = re.split(r'(\d.*)', op) - if len(splits) >= 2: - key, value = splits[:2] - options[key] = value - - # Check stride - assert (('s' in options and len(options['s']) == 1) or - (len(options['s']) == 2 and options['s'][0] == options['s'][1])) - - return BlockArgs( - kernel_size=int(options['k']), - num_repeat=int(options['r']), - input_filters=int(options['i']), - output_filters=int(options['o']), - expand_ratio=int(options['e']), - id_skip=('noskip' not in block_string), - se_ratio=float(options['se']) if 'se' in options else None, - stride=[int(options['s'][0])]) - - @staticmethod - def _encode_block_string(block): - """Encodes a block to a string.""" - args = [ - 'r%d' % block.num_repeat, - 'k%d' % block.kernel_size, - 's%d%d' % (block.strides[0], block.strides[1]), - 'e%s' % block.expand_ratio, - 'i%d' % block.input_filters, - 'o%d' % block.output_filters - ] - if 0 < block.se_ratio <= 1: - args.append('se%s' % block.se_ratio) - if block.id_skip is False: - args.append('noskip') - return '_'.join(args) - - @staticmethod - def decode(string_list): - """ - Decodes a list of string notations to specify blocks inside the network. - :param string_list: a list of strings, each string is a notation of block - :return: a list of BlockArgs namedtuples of block args - """ - assert isinstance(string_list, list) - blocks_args = [] - for block_string in string_list: - blocks_args.append(BlockDecoder._decode_block_string(block_string)) - return blocks_args - - @staticmethod - def encode(blocks_args): - """ - Encodes a list of BlockArgs to a list of strings. - :param blocks_args: a list of BlockArgs namedtuples of block args - :return: a list of strings, each string is a notation of block - """ - block_strings = [] - for block in blocks_args: - block_strings.append(BlockDecoder._encode_block_string(block)) - return block_strings - - -def EfficientNetB0_small(is_test=False, - padding_type='SAME', - override_params=None, - use_se=False): - model = EfficientNet( - name='b0', - is_test=is_test, - padding_type=padding_type, - override_params=override_params, - use_se=use_se) - return model - - -def EfficientNetB0(is_test=False, - padding_type='SAME', - override_params=None, - use_se=True): - model = EfficientNet( - name='b0', - is_test=is_test, - padding_type=padding_type, - override_params=override_params, - use_se=use_se) - return model - - -def EfficientNetB1(is_test=False, - padding_type='SAME', - override_params=None, - use_se=True): - model = EfficientNet( - name='b1', - is_test=is_test, - padding_type=padding_type, - override_params=override_params, - use_se=use_se) - return model - - -def EfficientNetB2(is_test=False, - padding_type='SAME', - override_params=None, - use_se=True): - model = EfficientNet( - name='b2', - is_test=is_test, - padding_type=padding_type, - override_params=override_params, - use_se=use_se) - return model - - -def EfficientNetB3(is_test=False, - padding_type='SAME', - override_params=None, - use_se=True): - model = EfficientNet( - name='b3', - is_test=is_test, - padding_type=padding_type, - override_params=override_params, - use_se=use_se) - return model - - -def EfficientNetB4(is_test=False, - padding_type='SAME', - override_params=None, - use_se=True): - model = EfficientNet( - name='b4', - is_test=is_test, - padding_type=padding_type, - override_params=override_params, - use_se=use_se) - return model - - -def EfficientNetB5(is_test=False, - padding_type='SAME', - override_params=None, - use_se=True): - model = EfficientNet( - name='b5', - is_test=is_test, - padding_type=padding_type, - override_params=override_params, - use_se=use_se) - return model - - -def EfficientNetB6(is_test=False, - padding_type='SAME', - override_params=None, - use_se=True): - model = EfficientNet( - name='b6', - is_test=is_test, - padding_type=padding_type, - override_params=override_params, - use_se=use_se) - return model - - -def EfficientNetB7(is_test=False, - padding_type='SAME', - override_params=None, - use_se=True): - model = EfficientNet( - name='b7', - is_test=is_test, - padding_type=padding_type, - override_params=override_params, - use_se=use_se) - return model diff --git a/hub_module/modules/image/classification/efficientnetb3_imagenet/label_list.txt b/hub_module/modules/image/classification/efficientnetb3_imagenet/label_list.txt deleted file mode 100644 index a509c007..00000000 --- a/hub_module/modules/image/classification/efficientnetb3_imagenet/label_list.txt +++ /dev/null @@ -1,1000 +0,0 @@ -tench, Tinca tinca -goldfish, Carassius auratus -great white shark, white shark, man-eater, man-eating shark, Carcharodon carcharias -tiger shark, Galeocerdo cuvieri -hammerhead, hammerhead shark -electric ray, crampfish, numbfish, torpedo -stingray -cock -hen -ostrich, Struthio camelus -brambling, Fringilla montifringilla -goldfinch, Carduelis carduelis -house finch, linnet, Carpodacus mexicanus -junco, snowbird -indigo bunting, indigo finch, indigo bird, Passerina cyanea -robin, American robin, Turdus migratorius -bulbul -jay -magpie -chickadee -water ouzel, dipper -kite -bald eagle, American eagle, Haliaeetus leucocephalus -vulture -great grey owl, great gray owl, Strix nebulosa -European fire salamander, Salamandra salamandra -common newt, Triturus vulgaris -eft -spotted salamander, Ambystoma maculatum -axolotl, mud puppy, Ambystoma mexicanum -bullfrog, Rana catesbeiana -tree frog, tree-frog -tailed frog, bell toad, ribbed toad, tailed toad, Ascaphus trui -loggerhead, loggerhead turtle, Caretta caretta -leatherback turtle, leatherback, leathery turtle, Dermochelys coriacea -mud turtle -terrapin -box turtle, box tortoise -banded gecko -common iguana, iguana, Iguana iguana -American chameleon, anole, Anolis carolinensis -whiptail, whiptail lizard -agama -frilled lizard, Chlamydosaurus kingi -alligator lizard -Gila monster, Heloderma suspectum -green lizard, Lacerta viridis -African chameleon, Chamaeleo chamaeleon -Komodo dragon, Komodo lizard, dragon lizard, giant lizard, Varanus komodoensis -African crocodile, Nile crocodile, Crocodylus niloticus -American alligator, Alligator mississipiensis -triceratops -thunder snake, worm snake, Carphophis amoenus -ringneck snake, ring-necked snake, ring snake -hognose snake, puff adder, sand viper -green snake, grass snake -king snake, kingsnake -garter snake, grass snake -water snake -vine snake -night snake, Hypsiglena torquata -boa constrictor, Constrictor constrictor -rock python, rock snake, Python sebae -Indian cobra, Naja naja -green mamba -sea snake -horned viper, cerastes, sand viper, horned asp, Cerastes cornutus -diamondback, diamondback rattlesnake, Crotalus adamanteus -sidewinder, horned rattlesnake, Crotalus cerastes -trilobite -harvestman, daddy longlegs, Phalangium opilio -scorpion -black and gold garden spider, Argiope aurantia -barn spider, Araneus cavaticus -garden spider, Aranea diademata -black widow, Latrodectus mactans -tarantula -wolf spider, hunting spider -tick -centipede -black grouse -ptarmigan -ruffed grouse, partridge, Bonasa umbellus -prairie chicken, prairie grouse, prairie fowl -peacock -quail -partridge -African grey, African gray, Psittacus erithacus -macaw -sulphur-crested cockatoo, Kakatoe galerita, Cacatua galerita -lorikeet -coucal -bee eater -hornbill -hummingbird -jacamar -toucan -drake -red-breasted merganser, Mergus serrator -goose -black swan, Cygnus atratus -tusker -echidna, spiny anteater, anteater -platypus, duckbill, duckbilled platypus, duck-billed platypus, Ornithorhynchus anatinus -wallaby, brush kangaroo -koala, koala bear, kangaroo bear, native bear, Phascolarctos cinereus -wombat -jellyfish -sea anemone, anemone -brain coral -flatworm, platyhelminth -nematode, nematode worm, roundworm -conch -snail -slug -sea slug, nudibranch -chiton, coat-of-mail shell, sea cradle, polyplacophore -chambered nautilus, pearly nautilus, nautilus -Dungeness crab, Cancer magister -rock crab, Cancer irroratus -fiddler crab -king crab, Alaska crab, Alaskan king crab, Alaska king crab, Paralithodes camtschatica -American lobster, Northern lobster, Maine lobster, Homarus americanus -spiny lobster, langouste, rock lobster, crawfish, crayfish, sea crawfish -crayfish, crawfish, crawdad, crawdaddy -hermit crab -isopod -white stork, Ciconia ciconia -black stork, Ciconia nigra -spoonbill -flamingo -little blue heron, Egretta caerulea -American egret, great white heron, Egretta albus -bittern -crane -limpkin, Aramus pictus -European gallinule, Porphyrio porphyrio -American coot, marsh hen, mud hen, water hen, Fulica americana -bustard -ruddy turnstone, Arenaria interpres -red-backed sandpiper, dunlin, Erolia alpina -redshank, Tringa totanus -dowitcher -oystercatcher, oyster catcher -pelican -king penguin, Aptenodytes patagonica -albatross, mollymawk -grey whale, gray whale, devilfish, Eschrichtius gibbosus, Eschrichtius robustus -killer whale, killer, orca, grampus, sea wolf, Orcinus orca -dugong, Dugong dugon -sea lion -Chihuahua -Japanese spaniel -Maltese dog, Maltese terrier, Maltese -Pekinese, Pekingese, Peke -Shih-Tzu -Blenheim spaniel -papillon -toy terrier -Rhodesian ridgeback -Afghan hound, Afghan -basset, basset hound -beagle -bloodhound, sleuthhound -bluetick -black-and-tan coonhound -Walker hound, Walker foxhound -English foxhound -redbone -borzoi, Russian wolfhound -Irish wolfhound -Italian greyhound -whippet -Ibizan hound, Ibizan Podenco -Norwegian elkhound, elkhound -otterhound, otter hound -Saluki, gazelle hound -Scottish deerhound, deerhound -Weimaraner -Staffordshire bullterrier, Staffordshire bull terrier -American Staffordshire terrier, Staffordshire terrier, American pit bull terrier, pit bull terrier -Bedlington terrier -Border terrier -Kerry blue terrier -Irish terrier -Norfolk terrier -Norwich terrier -Yorkshire terrier -wire-haired fox terrier -Lakeland terrier -Sealyham terrier, Sealyham -Airedale, Airedale terrier -cairn, cairn terrier -Australian terrier -Dandie Dinmont, Dandie Dinmont terrier -Boston bull, Boston terrier -miniature schnauzer -giant schnauzer -standard schnauzer -Scotch terrier, Scottish terrier, Scottie -Tibetan terrier, chrysanthemum dog -silky terrier, Sydney silky -soft-coated wheaten terrier -West Highland white terrier -Lhasa, Lhasa apso -flat-coated retriever -curly-coated retriever -golden retriever -Labrador retriever -Chesapeake Bay retriever -German short-haired pointer -vizsla, Hungarian pointer -English setter -Irish setter, red setter -Gordon setter -Brittany spaniel -clumber, clumber spaniel -English springer, English springer spaniel -Welsh springer spaniel -cocker spaniel, English cocker spaniel, cocker -Sussex spaniel -Irish water spaniel -kuvasz -schipperke -groenendael -malinois -briard -kelpie -komondor -Old English sheepdog, bobtail -Shetland sheepdog, Shetland sheep dog, Shetland -collie -Border collie -Bouvier des Flandres, Bouviers des Flandres -Rottweiler -German shepherd, German shepherd dog, German police dog, alsatian -Doberman, Doberman pinscher -miniature pinscher -Greater Swiss Mountain dog -Bernese mountain dog -Appenzeller -EntleBucher -boxer -bull mastiff -Tibetan mastiff -French bulldog -Great Dane -Saint Bernard, St Bernard -Eskimo dog, husky -malamute, malemute, Alaskan malamute -Siberian husky -dalmatian, coach dog, carriage dog -affenpinscher, monkey pinscher, monkey dog -basenji -pug, pug-dog -Leonberg -Newfoundland, Newfoundland dog -Great Pyrenees -Samoyed, Samoyede -Pomeranian -chow, chow chow -keeshond -Brabancon griffon -Pembroke, Pembroke Welsh corgi -Cardigan, Cardigan Welsh corgi -toy poodle -miniature poodle -standard poodle -Mexican hairless -timber wolf, grey wolf, gray wolf, Canis lupus -white wolf, Arctic wolf, Canis lupus tundrarum -red wolf, maned wolf, Canis rufus, Canis niger -coyote, prairie wolf, brush wolf, Canis latrans -dingo, warrigal, warragal, Canis dingo -dhole, Cuon alpinus -African hunting dog, hyena dog, Cape hunting dog, Lycaon pictus -hyena, hyaena -red fox, Vulpes vulpes -kit fox, Vulpes macrotis -Arctic fox, white fox, Alopex lagopus -grey fox, gray fox, Urocyon cinereoargenteus -tabby, tabby cat -tiger cat -Persian cat -Siamese cat, Siamese -Egyptian cat -cougar, puma, catamount, mountain lion, painter, panther, Felis concolor -lynx, catamount -leopard, Panthera pardus -snow leopard, ounce, Panthera uncia -jaguar, panther, Panthera onca, Felis onca -lion, king of beasts, Panthera leo -tiger, Panthera tigris -cheetah, chetah, Acinonyx jubatus -brown bear, bruin, Ursus arctos -American black bear, black bear, Ursus americanus, Euarctos americanus -ice bear, polar bear, Ursus Maritimus, Thalarctos maritimus -sloth bear, Melursus ursinus, Ursus ursinus -mongoose -meerkat, mierkat -tiger beetle -ladybug, ladybeetle, lady beetle, ladybird, ladybird beetle -ground beetle, carabid beetle -long-horned beetle, longicorn, longicorn beetle -leaf beetle, chrysomelid -dung beetle -rhinoceros beetle -weevil -fly -bee -ant, emmet, pismire -grasshopper, hopper -cricket -walking stick, walkingstick, stick insect -cockroach, roach -mantis, mantid -cicada, cicala -leafhopper -lacewing, lacewing fly -dragonfly, darning needle, devil's darning needle, sewing needle, snake feeder, snake doctor, mosquito hawk, skeeter hawk -damselfly -admiral -ringlet, ringlet butterfly -monarch, monarch butterfly, milkweed butterfly, Danaus plexippus -cabbage butterfly -sulphur butterfly, sulfur butterfly -lycaenid, lycaenid butterfly -starfish, sea star -sea urchin -sea cucumber, holothurian -wood rabbit, cottontail, cottontail rabbit -hare -Angora, Angora rabbit -hamster -porcupine, hedgehog -fox squirrel, eastern fox squirrel, Sciurus niger -marmot -beaver -guinea pig, Cavia cobaya -sorrel -zebra -hog, pig, grunter, squealer, Sus scrofa -wild boar, boar, Sus scrofa -warthog -hippopotamus, hippo, river horse, Hippopotamus amphibius -ox -water buffalo, water ox, Asiatic buffalo, Bubalus bubalis -bison -ram, tup -bighorn, bighorn sheep, cimarron, Rocky Mountain bighorn, Rocky Mountain sheep, Ovis canadensis -ibex, Capra ibex -hartebeest -impala, Aepyceros melampus -gazelle -Arabian camel, dromedary, Camelus dromedarius -llama -weasel -mink -polecat, fitch, foulmart, foumart, Mustela putorius -black-footed ferret, ferret, Mustela nigripes -otter -skunk, polecat, wood pussy -badger -armadillo -three-toed sloth, ai, Bradypus tridactylus -orangutan, orang, orangutang, Pongo pygmaeus -gorilla, Gorilla gorilla -chimpanzee, chimp, Pan troglodytes -gibbon, Hylobates lar -siamang, Hylobates syndactylus, Symphalangus syndactylus -guenon, guenon monkey -patas, hussar monkey, Erythrocebus patas -baboon -macaque -langur -colobus, colobus monkey -proboscis monkey, Nasalis larvatus -marmoset -capuchin, ringtail, Cebus capucinus -howler monkey, howler -titi, titi monkey -spider monkey, Ateles geoffroyi -squirrel monkey, Saimiri sciureus -Madagascar cat, ring-tailed lemur, Lemur catta -indri, indris, Indri indri, Indri brevicaudatus -Indian elephant, Elephas maximus -African elephant, Loxodonta africana -lesser panda, red panda, panda, bear cat, cat bear, Ailurus fulgens -giant panda, panda, panda bear, coon bear, Ailuropoda melanoleuca -barracouta, snoek -eel -coho, cohoe, coho salmon, blue jack, silver salmon, Oncorhynchus kisutch -rock beauty, Holocanthus tricolor -anemone fish -sturgeon -gar, garfish, garpike, billfish, Lepisosteus osseus -lionfish -puffer, pufferfish, blowfish, globefish -abacus -abaya -academic gown, academic robe, judge's robe -accordion, piano accordion, squeeze box -acoustic guitar -aircraft carrier, carrier, flattop, attack aircraft carrier -airliner -airship, dirigible -altar -ambulance -amphibian, amphibious vehicle -analog clock -apiary, bee house -apron -ashcan, trash can, garbage can, wastebin, ash bin, ash-bin, ashbin, dustbin, trash barrel, trash bin -assault rifle, assault gun -backpack, back pack, knapsack, packsack, rucksack, haversack -bakery, bakeshop, bakehouse -balance beam, beam -balloon -ballpoint, ballpoint pen, ballpen, Biro -Band Aid -banjo -bannister, banister, balustrade, balusters, handrail -barbell -barber chair -barbershop -barn -barometer -barrel, cask -barrow, garden cart, lawn cart, wheelbarrow -baseball -basketball -bassinet -bassoon -bathing cap, swimming cap -bath towel -bathtub, bathing tub, bath, tub -beach wagon, station wagon, wagon, estate car, beach waggon, station waggon, waggon -beacon, lighthouse, beacon light, pharos -beaker -bearskin, busby, shako -beer bottle -beer glass -bell cote, bell cot -bib -bicycle-built-for-two, tandem bicycle, tandem -bikini, two-piece -binder, ring-binder -binoculars, field glasses, opera glasses -birdhouse -boathouse -bobsled, bobsleigh, bob -bolo tie, bolo, bola tie, bola -bonnet, poke bonnet -bookcase -bookshop, bookstore, bookstall -bottlecap -bow -bow tie, bow-tie, bowtie -brass, memorial tablet, plaque -brassiere, bra, bandeau -breakwater, groin, groyne, mole, bulwark, seawall, jetty -breastplate, aegis, egis -broom -bucket, pail -buckle -bulletproof vest -bullet train, bullet -butcher shop, meat market -cab, hack, taxi, taxicab -caldron, cauldron -candle, taper, wax light -cannon -canoe -can opener, tin opener -cardigan -car mirror -carousel, carrousel, merry-go-round, roundabout, whirligig -carpenter's kit, tool kit -carton -car wheel -cash machine, cash dispenser, automated teller machine, automatic teller machine, automated teller, automatic teller, ATM -cassette -cassette player -castle -catamaran -CD player -cello, violoncello -cellular telephone, cellular phone, cellphone, cell, mobile phone -chain -chainlink fence -chain mail, ring mail, mail, chain armor, chain armour, ring armor, ring armour -chain saw, chainsaw -chest -chiffonier, commode -chime, bell, gong -china cabinet, china closet -Christmas stocking -church, church building -cinema, movie theater, movie theatre, movie house, picture palace -cleaver, meat cleaver, chopper -cliff dwelling -cloak -clog, geta, patten, sabot -cocktail shaker -coffee mug -coffeepot -coil, spiral, volute, whorl, helix -combination lock -computer keyboard, keypad -confectionery, confectionary, candy store -container ship, containership, container vessel -convertible -corkscrew, bottle screw -cornet, horn, trumpet, trump -cowboy boot -cowboy hat, ten-gallon hat -cradle -crane -crash helmet -crate -crib, cot -Crock Pot -croquet ball -crutch -cuirass -dam, dike, dyke -desk -desktop computer -dial telephone, dial phone -diaper, nappy, napkin -digital clock -digital watch -dining table, board -dishrag, dishcloth -dishwasher, dish washer, dishwashing machine -disk brake, disc brake -dock, dockage, docking facility -dogsled, dog sled, dog sleigh -dome -doormat, welcome mat -drilling platform, offshore rig -drum, membranophone, tympan -drumstick -dumbbell -Dutch oven -electric fan, blower -electric guitar -electric locomotive -entertainment center -envelope -espresso maker -face powder -feather boa, boa -file, file cabinet, filing cabinet -fireboat -fire engine, fire truck -fire screen, fireguard -flagpole, flagstaff -flute, transverse flute -folding chair -football helmet -forklift -fountain -fountain pen -four-poster -freight car -French horn, horn -frying pan, frypan, skillet -fur coat -garbage truck, dustcart -gasmask, respirator, gas helmet -gas pump, gasoline pump, petrol pump, island dispenser -goblet -go-kart -golf ball -golfcart, golf cart -gondola -gong, tam-tam -gown -grand piano, grand -greenhouse, nursery, glasshouse -grille, radiator grille -grocery store, grocery, food market, market -guillotine -hair slide -hair spray -half track -hammer -hamper -hand blower, blow dryer, blow drier, hair dryer, hair drier -hand-held computer, hand-held microcomputer -handkerchief, hankie, hanky, hankey -hard disc, hard disk, fixed disk -harmonica, mouth organ, harp, mouth harp -harp -harvester, reaper -hatchet -holster -home theater, home theatre -honeycomb -hook, claw -hoopskirt, crinoline -horizontal bar, high bar -horse cart, horse-cart -hourglass -iPod -iron, smoothing iron -jack-o'-lantern -jean, blue jean, denim -jeep, landrover -jersey, T-shirt, tee shirt -jigsaw puzzle -jinrikisha, ricksha, rickshaw -joystick -kimono -knee pad -knot -lab coat, laboratory coat -ladle -lampshade, lamp shade -laptop, laptop computer -lawn mower, mower -lens cap, lens cover -letter opener, paper knife, paperknife -library -lifeboat -lighter, light, igniter, ignitor -limousine, limo -liner, ocean liner -lipstick, lip rouge -Loafer -lotion -loudspeaker, speaker, speaker unit, loudspeaker system, speaker system -loupe, jeweler's loupe -lumbermill, sawmill -magnetic compass -mailbag, postbag -mailbox, letter box -maillot -maillot, tank suit -manhole cover -maraca -marimba, xylophone -mask -matchstick -maypole -maze, labyrinth -measuring cup -medicine chest, medicine cabinet -megalith, megalithic structure -microphone, mike -microwave, microwave oven -military uniform -milk can -minibus -miniskirt, mini -minivan -missile -mitten -mixing bowl -mobile home, manufactured home -Model T -modem -monastery -monitor -moped -mortar -mortarboard -mosque -mosquito net -motor scooter, scooter -mountain bike, all-terrain bike, off-roader -mountain tent -mouse, computer mouse -mousetrap -moving van -muzzle -nail -neck brace -necklace -nipple -notebook, notebook computer -obelisk -oboe, hautboy, hautbois -ocarina, sweet potato -odometer, hodometer, mileometer, milometer -oil filter -organ, pipe organ -oscilloscope, scope, cathode-ray oscilloscope, CRO -overskirt -oxcart -oxygen mask -packet -paddle, boat paddle -paddlewheel, paddle wheel -padlock -paintbrush -pajama, pyjama, pj's, jammies -palace -panpipe, pandean pipe, syrinx -paper towel -parachute, chute -parallel bars, bars -park bench -parking meter -passenger car, coach, carriage -patio, terrace -pay-phone, pay-station -pedestal, plinth, footstall -pencil box, pencil case -pencil sharpener -perfume, essence -Petri dish -photocopier -pick, plectrum, plectron -pickelhaube -picket fence, paling -pickup, pickup truck -pier -piggy bank, penny bank -pill bottle -pillow -ping-pong ball -pinwheel -pirate, pirate ship -pitcher, ewer -plane, carpenter's plane, woodworking plane -planetarium -plastic bag -plate rack -plow, plough -plunger, plumber's helper -Polaroid camera, Polaroid Land camera -pole -police van, police wagon, paddy wagon, patrol wagon, wagon, black Maria -poncho -pool table, billiard table, snooker table -pop bottle, soda bottle -pot, flowerpot -potter's wheel -power drill -prayer rug, prayer mat -printer -prison, prison house -projectile, missile -projector -puck, hockey puck -punching bag, punch bag, punching ball, punchball -purse -quill, quill pen -quilt, comforter, comfort, puff -racer, race car, racing car -racket, racquet -radiator -radio, wireless -radio telescope, radio reflector -rain barrel -recreational vehicle, RV, R.V. -reel -reflex camera -refrigerator, icebox -remote control, remote -restaurant, eating house, eating place, eatery -revolver, six-gun, six-shooter -rifle -rocking chair, rocker -rotisserie -rubber eraser, rubber, pencil eraser -rugby ball -rule, ruler -running shoe -safe -safety pin -saltshaker, salt shaker -sandal -sarong -sax, saxophone -scabbard -scale, weighing machine -school bus -schooner -scoreboard -screen, CRT screen -screw -screwdriver -seat belt, seatbelt -sewing machine -shield, buckler -shoe shop, shoe-shop, shoe store -shoji -shopping basket -shopping cart -shovel -shower cap -shower curtain -ski -ski mask -sleeping bag -slide rule, slipstick -sliding door -slot, one-armed bandit -snorkel -snowmobile -snowplow, snowplough -soap dispenser -soccer ball -sock -solar dish, solar collector, solar furnace -sombrero -soup bowl -space bar -space heater -space shuttle -spatula -speedboat -spider web, spider's web -spindle -sports car, sport car -spotlight, spot -stage -steam locomotive -steel arch bridge -steel drum -stethoscope -stole -stone wall -stopwatch, stop watch -stove -strainer -streetcar, tram, tramcar, trolley, trolley car -stretcher -studio couch, day bed -stupa, tope -submarine, pigboat, sub, U-boat -suit, suit of clothes -sundial -sunglass -sunglasses, dark glasses, shades -sunscreen, sunblock, sun blocker -suspension bridge -swab, swob, mop -sweatshirt -swimming trunks, bathing trunks -swing -switch, electric switch, electrical switch -syringe -table lamp -tank, army tank, armored combat vehicle, armoured combat vehicle -tape player -teapot -teddy, teddy bear -television, television system -tennis ball -thatch, thatched roof -theater curtain, theatre curtain -thimble -thresher, thrasher, threshing machine -throne -tile roof -toaster -tobacco shop, tobacconist shop, tobacconist -toilet seat -torch -totem pole -tow truck, tow car, wrecker -toyshop -tractor -trailer truck, tractor trailer, trucking rig, rig, articulated lorry, semi -tray -trench coat -tricycle, trike, velocipede -trimaran -tripod -triumphal arch -trolleybus, trolley coach, trackless trolley -trombone -tub, vat -turnstile -typewriter keyboard -umbrella -unicycle, monocycle -upright, upright piano -vacuum, vacuum cleaner -vase -vault -velvet -vending machine -vestment -viaduct -violin, fiddle -volleyball -waffle iron -wall clock -wallet, billfold, notecase, pocketbook -wardrobe, closet, press -warplane, military plane -washbasin, handbasin, washbowl, lavabo, wash-hand basin -washer, automatic washer, washing machine -water bottle -water jug -water tower -whiskey jug -whistle -wig -window screen -window shade -Windsor tie -wine bottle -wing -wok -wooden spoon -wool, woolen, woollen -worm fence, snake fence, snake-rail fence, Virginia fence -wreck -yawl -yurt -web site, website, internet site, site -comic book -crossword puzzle, crossword -street sign -traffic light, traffic signal, stoplight -book jacket, dust cover, dust jacket, dust wrapper -menu -plate -guacamole -consomme -hot pot, hotpot -trifle -ice cream, icecream -ice lolly, lolly, lollipop, popsicle -French loaf -bagel, beigel -pretzel -cheeseburger -hotdog, hot dog, red hot -mashed potato -head cabbage -broccoli -cauliflower -zucchini, courgette -spaghetti squash -acorn squash -butternut squash -cucumber, cuke -artichoke, globe artichoke -bell pepper -cardoon -mushroom -Granny Smith -strawberry -orange -lemon -fig -pineapple, ananas -banana -jackfruit, jak, jack -custard apple -pomegranate -hay -carbonara -chocolate sauce, chocolate syrup -dough -meat loaf, meatloaf -pizza, pizza pie -potpie -burrito -red wine -espresso -cup -eggnog -alp -bubble -cliff, drop, drop-off -coral reef -geyser -lakeside, lakeshore -promontory, headland, head, foreland -sandbar, sand bar -seashore, coast, seacoast, sea-coast -valley, vale -volcano -ballplayer, baseball player -groom, bridegroom -scuba diver -rapeseed -daisy -yellow lady's slipper, yellow lady-slipper, Cypripedium calceolus, Cypripedium parviflorum -corn -acorn -hip, rose hip, rosehip -buckeye, horse chestnut, conker -coral fungus -agaric -gyromitra -stinkhorn, carrion fungus -earthstar -hen-of-the-woods, hen of the woods, Polyporus frondosus, Grifola frondosa -bolete -ear, spike, capitulum -toilet tissue, toilet paper, bathroom tissue diff --git a/hub_module/modules/image/classification/efficientnetb3_imagenet/layers.py b/hub_module/modules/image/classification/efficientnetb3_imagenet/layers.py deleted file mode 100644 index 00f792b5..00000000 --- a/hub_module/modules/image/classification/efficientnetb3_imagenet/layers.py +++ /dev/null @@ -1,248 +0,0 @@ -# -*- coding:utf-8 -*- -# copyright (c) 2020 PaddlePaddle Authors. All Rights Reserve. -# -# Licensed under the Apache License, Version 2.0 (the "License"); -# you may not use this file except in compliance with the License. -# You may obtain a copy of the License at -# -# http://www.apache.org/licenses/LICENSE-2.0 -# -# Unless required by applicable law or agreed to in writing, software -# distributed under the License is distributed on an "AS IS" BASIS, -# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. -# See the License for the specific language governing permissions and -# limitations under the License. - -from __future__ import absolute_import -from __future__ import division -from __future__ import print_function - -import math -import warnings - -import paddle.fluid as fluid - - -def initial_type(name, - input, - op_type, - fan_out, - init="google", - use_bias=False, - filter_size=0, - stddev=0.02): - if init == "kaiming": - if op_type == 'conv': - fan_in = input.shape[1] * filter_size * filter_size - elif op_type == 'deconv': - fan_in = fan_out * filter_size * filter_size - else: - if len(input.shape) > 2: - fan_in = input.shape[1] * input.shape[2] * input.shape[3] - else: - fan_in = input.shape[1] - bound = 1 / math.sqrt(fan_in) - param_attr = fluid.ParamAttr( - name=name + "_weights", - initializer=fluid.initializer.Uniform(low=-bound, high=bound)) - if use_bias == True: - bias_attr = fluid.ParamAttr( - name=name + '_offset', - initializer=fluid.initializer.Uniform(low=-bound, high=bound)) - else: - bias_attr = False - elif init == 'google': - n = filter_size * filter_size * fan_out - param_attr = fluid.ParamAttr( - name=name + "_weights", - initializer=fluid.initializer.NormalInitializer( - loc=0.0, scale=math.sqrt(2.0 / n))) - if use_bias == True: - bias_attr = fluid.ParamAttr( - name=name + "_offset", - initializer=fluid.initializer.Constant(0.0)) - else: - bias_attr = False - - else: - param_attr = fluid.ParamAttr( - name=name + "_weights", - initializer=fluid.initializer.NormalInitializer( - loc=0.0, scale=stddev)) - if use_bias == True: - bias_attr = fluid.ParamAttr( - name=name + "_offset", - initializer=fluid.initializer.Constant(0.0)) - else: - bias_attr = False - return param_attr, bias_attr - - -def cal_padding(img_size, stride, filter_size, dilation=1): - """Calculate padding size.""" - if img_size % stride == 0: - out_size = max(filter_size - stride, 0) - else: - out_size = max(filter_size - (img_size % stride), 0) - return out_size // 2, out_size - out_size // 2 - - -def init_batch_norm_layer(name="batch_norm"): - param_attr = fluid.ParamAttr( - name=name + '_scale', initializer=fluid.initializer.Constant(1.0)) - bias_attr = fluid.ParamAttr( - name=name + '_offset', - initializer=fluid.initializer.Constant(value=0.0)) - return param_attr, bias_attr - - -def init_fc_layer(fout, name='fc'): - n = fout # fan-out - init_range = 1.0 / math.sqrt(n) - - param_attr = fluid.ParamAttr( - name=name + '_weights', - initializer=fluid.initializer.UniformInitializer( - low=-init_range, high=init_range)) - bias_attr = fluid.ParamAttr( - name=name + '_offset', - initializer=fluid.initializer.Constant(value=0.0)) - return param_attr, bias_attr - - -def norm_layer(input, norm_type='batch_norm', name=None): - if norm_type == 'batch_norm': - param_attr = fluid.ParamAttr( - name=name + '_weights', initializer=fluid.initializer.Constant(1.0)) - bias_attr = fluid.ParamAttr( - name=name + '_offset', - initializer=fluid.initializer.Constant(value=0.0)) - return fluid.layers.batch_norm( - input, - param_attr=param_attr, - bias_attr=bias_attr, - moving_mean_name=name + '_mean', - moving_variance_name=name + '_variance') - - elif norm_type == 'instance_norm': - helper = fluid.layer_helper.LayerHelper("instance_norm", **locals()) - dtype = helper.input_dtype() - epsilon = 1e-5 - mean = fluid.layers.reduce_mean(input, dim=[2, 3], keep_dim=True) - var = fluid.layers.reduce_mean( - fluid.layers.square(input - mean), dim=[2, 3], keep_dim=True) - if name is not None: - scale_name = name + "_scale" - offset_name = name + "_offset" - scale_param = fluid.ParamAttr( - name=scale_name, - initializer=fluid.initializer.Constant(1.0), - trainable=True) - offset_param = fluid.ParamAttr( - name=offset_name, - initializer=fluid.initializer.Constant(0.0), - trainable=True) - scale = helper.create_parameter( - attr=scale_param, shape=input.shape[1:2], dtype=dtype) - offset = helper.create_parameter( - attr=offset_param, shape=input.shape[1:2], dtype=dtype) - - tmp = fluid.layers.elementwise_mul(x=(input - mean), y=scale, axis=1) - tmp = tmp / fluid.layers.sqrt(var + epsilon) - tmp = fluid.layers.elementwise_add(tmp, offset, axis=1) - return tmp - else: - raise NotImplementedError("norm tyoe: [%s] is not support" % norm_type) - - -def conv2d(input, - num_filters=64, - filter_size=7, - stride=1, - stddev=0.02, - padding=0, - groups=None, - name="conv2d", - norm=None, - act=None, - relufactor=0.0, - use_bias=False, - padding_type=None, - initial="normal", - use_cudnn=True): - - if padding != 0 and padding_type != None: - warnings.warn( - 'padding value and padding type are set in the same time, and the final padding width and padding height are computed by padding_type' - ) - - param_attr, bias_attr = initial_type( - name=name, - input=input, - op_type='conv', - fan_out=num_filters, - init=initial, - use_bias=use_bias, - filter_size=filter_size, - stddev=stddev) - - def get_padding(filter_size, stride=1, dilation=1): - padding = ((stride - 1) + dilation * (filter_size - 1)) // 2 - return padding - - need_crop = False - if padding_type == "SAME": - top_padding, bottom_padding = cal_padding(input.shape[2], stride, - filter_size) - left_padding, right_padding = cal_padding(input.shape[2], stride, - filter_size) - height_padding = bottom_padding - width_padding = right_padding - if top_padding != bottom_padding or left_padding != right_padding: - height_padding = top_padding + stride - width_padding = left_padding + stride - need_crop = True - padding = [height_padding, width_padding] - elif padding_type == "VALID": - height_padding = 0 - width_padding = 0 - padding = [height_padding, width_padding] - elif padding_type == "DYNAMIC": - padding = get_padding(filter_size, stride) - else: - padding = padding - - conv = fluid.layers.conv2d( - input, - num_filters, - filter_size, - groups=groups, - name=name, - stride=stride, - padding=padding, - use_cudnn=use_cudnn, - param_attr=param_attr, - bias_attr=bias_attr) - - if need_crop: - conv = conv[:, :, 1:, 1:] - - if norm is not None: - conv = norm_layer(input=conv, norm_type=norm, name=name + "_norm") - if act == 'relu': - conv = fluid.layers.relu(conv, name=name + '_relu') - elif act == 'leaky_relu': - conv = fluid.layers.leaky_relu( - conv, alpha=relufactor, name=name + '_leaky_relu') - elif act == 'tanh': - conv = fluid.layers.tanh(conv, name=name + '_tanh') - elif act == 'sigmoid': - conv = fluid.layers.sigmoid(conv, name=name + '_sigmoid') - elif act == 'swish': - conv = fluid.layers.swish(conv, name=name + '_swish') - elif act == None: - conv = conv - else: - raise NotImplementedError("activation: [%s] is not support" % act) - - return conv diff --git a/hub_module/modules/image/classification/efficientnetb3_imagenet/module.py b/hub_module/modules/image/classification/efficientnetb3_imagenet/module.py index 98606568..24759985 100644 --- a/hub_module/modules/image/classification/efficientnetb3_imagenet/module.py +++ b/hub_module/modules/image/classification/efficientnetb3_imagenet/module.py @@ -1,4 +1,3 @@ -# -*- coding:utf-8 -*- # copyright (c) 2020 PaddlePaddle Authors. All Rights Reserve. # # Licensed under the Apache License, Version 2.0 (the "License"); @@ -13,318 +12,747 @@ # See the License for the specific language governing permissions and # limitations under the License. -from __future__ import absolute_import -from __future__ import division - -import ast -import argparse import os - -import numpy as np -import paddle.fluid as fluid -import paddlehub as hub -from paddle.fluid.core import PaddleTensor, AnalysisConfig, create_paddle_predictor -from paddlehub.module.module import moduleinfo, runnable, serving -from paddlehub.common.paddle_helper import add_vars_prefix - -from efficientnetb3_imagenet.processor import postprocess, base64_to_cv2 -from efficientnetb3_imagenet.data_feed import reader -from efficientnetb3_imagenet.efficientnet import EfficientNetB3 - - -@moduleinfo( - name="efficientnetb3_imagenet", - type="CV/image_classification", - author="paddlepaddle", - author_email="paddle-dev@baidu.com", - summary= - "EfficientNetB3 is a image classfication model, this module is trained with imagenet datasets.", - version="1.1.0") -class EfficientNetB3ImageNet(hub.Module): - def _initialize(self): - self.default_pretrained_model_path = os.path.join( - self.directory, "efficientnetb3_imagenet_infer_model") - label_file = os.path.join(self.directory, "label_list.txt") - with open(label_file, 'r', encoding='utf-8') as file: - self.label_list = file.read().split("\n")[:-1] - self.classification = self.classify - self._set_config() - - def get_expected_image_width(self): - return 224 - - def get_expected_image_height(self): - return 224 - - def get_pretrained_images_mean(self): - im_mean = np.array([0.485, 0.456, 0.406]).reshape(1, 3) - return im_mean - - def get_pretrained_images_std(self): - im_std = np.array([0.229, 0.224, 0.225]).reshape(1, 3) - return im_std - - def _set_config(self): - """ - predictor config setting - """ - cpu_config = AnalysisConfig(self.default_pretrained_model_path) - cpu_config.disable_glog_info() - cpu_config.disable_gpu() - self.cpu_predictor = create_paddle_predictor(cpu_config) - - try: - _places = os.environ["CUDA_VISIBLE_DEVICES"] - int(_places[0]) - use_gpu = True - except: - use_gpu = False - if use_gpu: - gpu_config = AnalysisConfig(self.default_pretrained_model_path) - gpu_config.disable_glog_info() - gpu_config.enable_use_gpu( - memory_pool_init_size_mb=1000, device_id=0) - self.gpu_predictor = create_paddle_predictor(gpu_config) - - def context(self, - trainable=True, - pretrained=True, - override_params=None, - phase='train'): - """context for transfer learning. - - Args: - trainable (bool): Set parameters in program to be trainable. - pretrained (bool) : Whether to load pretrained model. - - Returns: - inputs (dict): key is 'image', corresponding vaule is image tensor. - outputs (dict): key is : - 'classification', corresponding value is the result of classification. - 'feature_map', corresponding value is the result of the layer before the fully connected layer. - context_prog (fluid.Program): program for transfer learning. - """ - if phase in ["dev", "test", "predict", "eval"]: - is_test = True - elif phase in ["train"]: - is_test = False - else: - raise ValueError( - "Phase %s is error, which must be one of train, dev, test, eval and predict." - % phase) - - context_prog = fluid.Program() - startup_prog = fluid.Program() - with fluid.program_guard(context_prog, startup_prog): - with fluid.unique_name.guard(): - image = fluid.layers.data( - name="image", shape=[3, 224, 224], dtype="float32") - efficientnet_b3 = EfficientNetB3( - override_params=override_params) - output, feature_map = efficientnet_b3.net( - input=image, - class_dim=len(self.label_list), - is_test=is_test) - - name_prefix = '@HUB_{}@'.format(self.name) - inputs = {'image': name_prefix + image.name} - outputs = { - 'classification': name_prefix + output.name, - 'feature_map': name_prefix + feature_map.name - } - add_vars_prefix(context_prog, name_prefix) - add_vars_prefix(startup_prog, name_prefix) - global_vars = context_prog.global_block().vars - inputs = { - key: global_vars[value] - for key, value in inputs.items() - } - outputs = { - key: global_vars[value] - for key, value in outputs.items() - } - - place = fluid.CPUPlace() - exe = fluid.Executor(place) - # pretrained - if pretrained: - - def _if_exist(var): - b = os.path.exists( - os.path.join(self.default_pretrained_model_path, - var.name)) - return b - - fluid.io.load_vars( - exe, - self.default_pretrained_model_path, - context_prog, - predicate=_if_exist) - else: - exe.run(startup_prog) - # trainable - for param in context_prog.global_block().iter_parameters(): - param.trainable = trainable - return inputs, outputs, context_prog - - def classify(self, - images=None, - paths=None, - batch_size=1, - use_gpu=False, - top_k=1): +import math +import collections +import re +import copy + +import paddle +from paddle import ParamAttr +import paddle.nn as nn +import paddle.nn.functional as F +from paddle.nn import Conv2d, BatchNorm, Linear, Dropout +from paddle.nn import AdaptiveAvgPool2d, MaxPool2d, AvgPool2d +from paddlehub.module.module import moduleinfo +from paddlehub.module.cv_module import ImageClassifierModule + +GlobalParams = collections.namedtuple('GlobalParams', [ + 'batch_norm_momentum', + 'batch_norm_epsilon', + 'dropout_rate', + 'num_classes', + 'width_coefficient', + 'depth_coefficient', + 'depth_divisor', + 'min_depth', + 'drop_connect_rate', +]) + +BlockArgs = collections.namedtuple( + 'BlockArgs', + ['kernel_size', 'num_repeat', 'input_filters', 'output_filters', 'expand_ratio', 'id_skip', 'stride', 'se_ratio']) + +GlobalParams.__new__.__defaults__ = (None, ) * len(GlobalParams._fields) +BlockArgs.__new__.__defaults__ = (None, ) * len(BlockArgs._fields) + + +def efficientnet_params(model_name: str): + """ Map EfficientNet model name to parameter coefficients. """ + params_dict = { + # Coefficients: width,depth,resolution,dropout + 'efficientnet-b3': (1.2, 1.4, 300, 0.3) + } + return params_dict[model_name] + + +def efficientnet(width_coefficient: float = None, + depth_coefficient: float = None, + dropout_rate: float = 0.2, + drop_connect_rate: float = 0.2): + """ Get block arguments according to parameter and coefficients. """ + blocks_args = [ + 'r1_k3_s11_e1_i32_o16_se0.25', + 'r2_k3_s22_e6_i16_o24_se0.25', + 'r2_k5_s22_e6_i24_o40_se0.25', + 'r3_k3_s22_e6_i40_o80_se0.25', + 'r3_k5_s11_e6_i80_o112_se0.25', + 'r4_k5_s22_e6_i112_o192_se0.25', + 'r1_k3_s11_e6_i192_o320_se0.25', + ] + blocks_args = BlockDecoder.decode(blocks_args) + + global_params = GlobalParams(batch_norm_momentum=0.99, + batch_norm_epsilon=1e-3, + dropout_rate=dropout_rate, + drop_connect_rate=drop_connect_rate, + num_classes=1000, + width_coefficient=width_coefficient, + depth_coefficient=depth_coefficient, + depth_divisor=8, + min_depth=None) + + return blocks_args, global_params + + +def get_model_params(model_name: str, override_params: dict): + """ Get the block args and global params for a given model """ + if model_name.startswith('efficientnet'): + w, d, _, p = efficientnet_params(model_name) + blocks_args, global_params = efficientnet(width_coefficient=w, depth_coefficient=d, dropout_rate=p) + else: + raise NotImplementedError('model name is not pre-defined: %s' % model_name) + if override_params: + global_params = global_params._replace(**override_params) + return blocks_args, global_params + + +def round_filters(filters: int, global_params: dict): + """ Calculate and round number of filters based on depth multiplier. """ + multiplier = global_params.width_coefficient + if not multiplier: + return filters + divisor = global_params.depth_divisor + min_depth = global_params.min_depth + filters *= multiplier + min_depth = min_depth or divisor + new_filters = max(min_depth, int(filters + divisor / 2) // divisor * divisor) + if new_filters < 0.9 * filters: # prevent rounding by more than 10% + new_filters += divisor + return int(new_filters) + + +def round_repeats(repeats: int, global_params: dict): + """ Round number of filters based on depth multiplier. """ + multiplier = global_params.depth_coefficient + if not multiplier: + return repeats + return int(math.ceil(multiplier * repeats)) + + +class BlockDecoder(object): + """ + Block Decoder, straight from the official TensorFlow repository. + """ + @staticmethod + def _decode_block_string(block_string: str): + """ Gets a block through a string notation of arguments. """ + assert isinstance(block_string, str) + + ops = block_string.split('_') + options = {} + for op in ops: + splits = re.split(r'(\d.*)', op) + if len(splits) >= 2: + key, value = splits[:2] + options[key] = value + + # Check stride + cond_1 = ('s' in options and len(options['s']) == 1) + cond_2 = ((len(options['s']) == 2) and (options['s'][0] == options['s'][1])) + assert (cond_1 or cond_2) + + return BlockArgs(kernel_size=int(options['k']), + num_repeat=int(options['r']), + input_filters=int(options['i']), + output_filters=int(options['o']), + expand_ratio=int(options['e']), + id_skip=('noskip' not in block_string), + se_ratio=float(options['se']) if 'se' in options else None, + stride=[int(options['s'][0])]) + + @staticmethod + def _encode_block_string(block): + """Encodes a block to a string.""" + args = [ + 'r%d' % block.num_repeat, + 'k%d' % block.kernel_size, + 's%d%d' % (block.strides[0], block.strides[1]), + 'e%s' % block.expand_ratio, + 'i%d' % block.input_filters, + 'o%d' % block.output_filters + ] + if 0 < block.se_ratio <= 1: + args.append('se%s' % block.se_ratio) + if block.id_skip is False: + args.append('noskip') + return '_'.join(args) + + @staticmethod + def decode(string_list: list): """ - API for image classification. - - Args: - images (list[numpy.ndarray]): data of images, shape of each is [H, W, C], color space must be BGR. - paths (list[str]): The paths of images. - batch_size (int): batch size. - use_gpu (bool): Whether to use gpu. - top_k (int): Return top k results. + Decode a list of string notations to specify blocks in the network. - Returns: - res (list[dict]): The classfication results. + string_list: list of strings, each string is a notation of block + return + list of BlockArgs namedtuples of block args """ - if use_gpu: - try: - _places = os.environ["CUDA_VISIBLE_DEVICES"] - int(_places[0]) - except: - raise RuntimeError( - "Environment Variable CUDA_VISIBLE_DEVICES is not set correctly. If you wanna use gpu, please set CUDA_VISIBLE_DEVICES as cuda_device_id." - ) - - all_data = list() - for yield_data in reader(images, paths): - all_data.append(yield_data) - - total_num = len(all_data) - loop_num = int(np.ceil(total_num / batch_size)) - - res = list() - for iter_id in range(loop_num): - batch_data = list() - handle_id = iter_id * batch_size - for image_id in range(batch_size): - try: - batch_data.append(all_data[handle_id + image_id]) - except: - pass - # feed batch image - batch_image = np.array([data['image'] for data in batch_data]) - batch_image = PaddleTensor(batch_image.copy()) - predictor_output = self.gpu_predictor.run([ - batch_image - ]) if use_gpu else self.cpu_predictor.run([batch_image]) - out = postprocess( - data_out=predictor_output[0].as_ndarray(), - label_list=self.label_list, - top_k=top_k) - res += out - return res - - def save_inference_model(self, - dirname, - model_filename=None, - params_filename=None, - combined=True): - if combined: - model_filename = "__model__" if not model_filename else model_filename - params_filename = "__params__" if not params_filename else params_filename - place = fluid.CPUPlace() - exe = fluid.Executor(place) - - program, feeded_var_names, target_vars = fluid.io.load_inference_model( - dirname=self.default_pretrained_model_path, executor=exe) - - fluid.io.save_inference_model( - dirname=dirname, - main_program=program, - executor=exe, - feeded_var_names=feeded_var_names, - target_vars=target_vars, - model_filename=model_filename, - params_filename=params_filename) - - @serving - def serving_method(self, images, **kwargs): + assert isinstance(string_list, list) + blocks_args = [] + for block_string in string_list: + blocks_args.append(BlockDecoder._decode_block_string(block_string)) + return blocks_args + + @staticmethod + def encode(blocks_args: list): """ - Run as a service. - """ - images_decode = [base64_to_cv2(image) for image in images] - results = self.classify(images=images_decode, **kwargs) - return results + Encodes a list of BlockArgs to a list of strings. - @runnable - def run_cmd(self, argvs): - """ - Run as a command. + :param blocks_args: a list of BlockArgs namedtuples of block args + :return: a list of strings, each string is a notation of block """ - self.parser = argparse.ArgumentParser( - description="Run the {} module.".format(self.name), - prog='hub run {}'.format(self.name), - usage='%(prog)s', - add_help=True) - self.arg_input_group = self.parser.add_argument_group( - title="Input options", description="Input data. Required") - self.arg_config_group = self.parser.add_argument_group( - title="Config options", - description= - "Run configuration for controlling module behavior, not required.") - self.add_module_config_arg() - self.add_module_input_arg() - args = self.parser.parse_args(argvs) - results = self.classify( - paths=[args.input_path], - batch_size=args.batch_size, - use_gpu=args.use_gpu) - return results - - def add_module_config_arg(self): - """ - Add the command config options. - """ - self.arg_config_group.add_argument( - '--use_gpu', - type=ast.literal_eval, - default=False, - help="whether use GPU or not.") - self.arg_config_group.add_argument( - '--batch_size', - type=ast.literal_eval, - default=1, - help="batch size.") - self.arg_config_group.add_argument( - '--top_k', - type=ast.literal_eval, - default=1, - help="Return top k results.") - - def add_module_input_arg(self): - """ - Add the command input options. - """ - self.arg_input_group.add_argument( - '--input_path', type=str, help="path to image.") - - -if __name__ == '__main__': - b3 = EfficientNetB3ImageNet() - b3.context() - import cv2 - test_image = [ - cv2.imread( - '/mnt/zhangxuefei/program-paddle/PaddleHub/hub_module/tests/image_dataset/classification/animals/dog.jpeg' - ) - ] - res = b3.classification(images=test_image) - print(res) - res = b3.classification(paths=[ - '/mnt/zhangxuefei/program-paddle/PaddleHub/hub_module/tests/image_dataset/classification/animals/dog.jpeg' - ]) - print(res) - res = b3.classification(images=test_image) - print(res) - res = b3.classify(images=test_image) - print(res) + block_strings = [] + for block in blocks_args: + block_strings.append(BlockDecoder._encode_block_string(block)) + return block_strings + + +def initial_type(name: str, use_bias: bool = False): + param_attr = ParamAttr(name=name + "_weights") + if use_bias: + bias_attr = ParamAttr(name=name + "_offset") + else: + bias_attr = False + return param_attr, bias_attr + + +def init_batch_norm_layer(name: str = "batch_norm"): + param_attr = ParamAttr(name=name + "_scale") + bias_attr = ParamAttr(name=name + "_offset") + return param_attr, bias_attr + + +def init_fc_layer(name: str = "fc"): + param_attr = ParamAttr(name=name + "_weights") + bias_attr = ParamAttr(name=name + "_offset") + return param_attr, bias_attr + + +def cal_padding(img_size: int, stride: int, filter_size: int, dilation: int = 1): + """Calculate padding size.""" + if img_size % stride == 0: + out_size = max(filter_size - stride, 0) + else: + out_size = max(filter_size - (img_size % stride), 0) + return out_size // 2, out_size - out_size // 2 + + +inp_shape = {"b3": [300, 150, 150, 75, 38, 19, 19, 10]} + + +def _drop_connect(inputs: paddle.Tensor, prob: float, is_test: bool): + """Drop input connection""" + if is_test: + return inputs + keep_prob = 1.0 - prob + inputs_shape = paddle.shape(inputs) + random_tensor = keep_prob + paddle.rand(shape=[inputs_shape[0], 1, 1, 1]) + binary_tensor = paddle.floor(random_tensor) + output = inputs / keep_prob * binary_tensor + return output + + +class Conv2ds(nn.Layer): + """Basic conv layer""" + def __init__(self, + input_channels: int, + output_channels: int, + filter_size: int, + stride: int = 1, + padding: int = 0, + groups: int = None, + name: str = "conv2d", + act: str = None, + use_bias: bool = False, + padding_type: str = None, + model_name: str = None, + cur_stage: str = None): + super(Conv2ds, self).__init__() + assert act in [None, "swish", "sigmoid"] + self.act = act + + param_attr, bias_attr = initial_type(name=name, use_bias=use_bias) + + def get_padding(filter_size, stride=1, dilation=1): + padding = ((stride - 1) + dilation * (filter_size - 1)) // 2 + return padding + + inps = 1 if model_name == None and cur_stage == None else inp_shape[model_name][cur_stage] + self.need_crop = False + if padding_type == "SAME": + top_padding, bottom_padding = cal_padding(inps, stride, filter_size) + left_padding, right_padding = cal_padding(inps, stride, filter_size) + height_padding = bottom_padding + width_padding = right_padding + if top_padding != bottom_padding or left_padding != right_padding: + height_padding = top_padding + stride + width_padding = left_padding + stride + self.need_crop = True + padding = [height_padding, width_padding] + elif padding_type == "VALID": + height_padding = 0 + width_padding = 0 + padding = [height_padding, width_padding] + elif padding_type == "DYNAMIC": + padding = get_padding(filter_size, stride) + else: + padding = padding_type + + groups = 1 if groups is None else groups + self._conv = Conv2d(input_channels, + output_channels, + filter_size, + groups=groups, + stride=stride, + padding=padding, + weight_attr=param_attr, + bias_attr=bias_attr) + + def forward(self, inputs: paddle.Tensor): + x = self._conv(inputs) + if self.act == "swish": + x = F.swish(x) + elif self.act == "sigmoid": + x = F.sigmoid(x) + + if self.need_crop: + x = x[:, :, 1:, 1:] + return x + + +class ConvBNLayer(nn.Layer): + """Basic conv bn layer.""" + def __init__(self, + input_channels: int, + filter_size: int, + output_channels: int, + stride: int = 1, + num_groups: int = 1, + padding_type: str = "SAME", + conv_act: str = None, + bn_act: str = "swish", + use_bn: bool = True, + use_bias: bool = False, + name: str = None, + conv_name: str = None, + bn_name: str = None, + model_name: str = None, + cur_stage: str = None): + super(ConvBNLayer, self).__init__() + + self._conv = Conv2ds(input_channels=input_channels, + output_channels=output_channels, + filter_size=filter_size, + stride=stride, + groups=num_groups, + act=conv_act, + padding_type=padding_type, + name=conv_name, + use_bias=use_bias, + model_name=model_name, + cur_stage=cur_stage) + self.use_bn = use_bn + if use_bn is True: + bn_name = name + bn_name + param_attr, bias_attr = init_batch_norm_layer(bn_name) + + self._bn = BatchNorm(num_channels=output_channels, + act=bn_act, + momentum=0.99, + epsilon=0.001, + moving_mean_name=bn_name + "_mean", + moving_variance_name=bn_name + "_variance", + param_attr=param_attr, + bias_attr=bias_attr) + + def forward(self, inputs: paddle.Tensor): + if self.use_bn: + x = self._conv(inputs) + x = self._bn(x) + return x + else: + return self._conv(inputs) + + +class ExpandConvNorm(nn.Layer): + """Expand conv norm layer.""" + def __init__(self, + input_channels: int, + block_args: dict, + padding_type: str, + name: str = None, + model_name: str = None, + cur_stage: str = None): + super(ExpandConvNorm, self).__init__() + + self.oup = block_args.input_filters * block_args.expand_ratio + self.expand_ratio = block_args.expand_ratio + + if self.expand_ratio != 1: + self._conv = ConvBNLayer(input_channels, + 1, + self.oup, + bn_act=None, + padding_type=padding_type, + name=name, + conv_name=name + "_expand_conv", + bn_name="_bn0", + model_name=model_name, + cur_stage=cur_stage) + + def forward(self, inputs: paddle.Tensor): + if self.expand_ratio != 1: + return self._conv(inputs) + else: + return inputs + + +class DepthwiseConvNorm(nn.Layer): + """Depthwise conv norm layer.""" + def __init__(self, + input_channels: int, + block_args: dict, + padding_type: str, + name: str = None, + model_name: str = None, + cur_stage: str = None): + super(DepthwiseConvNorm, self).__init__() + + self.k = block_args.kernel_size + self.s = block_args.stride + if isinstance(self.s, list) or isinstance(self.s, tuple): + self.s = self.s[0] + oup = block_args.input_filters * block_args.expand_ratio + + self._conv = ConvBNLayer(input_channels, + self.k, + oup, + self.s, + num_groups=input_channels, + bn_act=None, + padding_type=padding_type, + name=name, + conv_name=name + "_depthwise_conv", + bn_name="_bn1", + model_name=model_name, + cur_stage=cur_stage) + + def forward(self, inputs: paddle.Tensor): + return self._conv(inputs) + + +class ProjectConvNorm(nn.Layer): + """Projection conv bn layer.""" + def __init__(self, + input_channels: int, + block_args: dict, + padding_type: str, + name: str = None, + model_name: str = None, + cur_stage: str = None): + super(ProjectConvNorm, self).__init__() + + final_oup = block_args.output_filters + + self._conv = ConvBNLayer(input_channels, + 1, + final_oup, + bn_act=None, + padding_type=padding_type, + name=name, + conv_name=name + "_project_conv", + bn_name="_bn2", + model_name=model_name, + cur_stage=cur_stage) + + def forward(self, inputs: paddle.Tensor): + return self._conv(inputs) + + +class SEBlock(nn.Layer): + """Basic Squeeze-and-Excitation block for Efficientnet.""" + def __init__(self, + input_channels: int, + num_squeezed_channels: int, + oup: int, + padding_type: str, + name: str = None, + model_name: str = None, + cur_stage: str = None): + super(SEBlock, self).__init__() + + self._pool = AdaptiveAvgPool2d(1) + self._conv1 = Conv2ds(input_channels, + num_squeezed_channels, + 1, + use_bias=True, + padding_type=padding_type, + act="swish", + name=name + "_se_reduce") + + self._conv2 = Conv2ds(num_squeezed_channels, + oup, + 1, + act="sigmoid", + use_bias=True, + padding_type=padding_type, + name=name + "_se_expand") + + def forward(self, inputs: paddle.Tensor): + x = self._pool(inputs) + x = self._conv1(x) + x = self._conv2(x) + return paddle.multiply(inputs, x) + + +class MbConvBlock(nn.Layer): + """Mobile inverted bottleneck convolution for Efficientnet.""" + def __init__(self, + input_channels: int, + block_args: dict, + padding_type: str, + use_se: bool, + name: str = None, + drop_connect_rate: float = None, + is_test: bool = False, + model_name: str = None, + cur_stage: str = None): + super(MbConvBlock, self).__init__() + + oup = block_args.input_filters * block_args.expand_ratio + self.block_args = block_args + self.has_se = use_se and (block_args.se_ratio is not None) and (0 < block_args.se_ratio <= 1) + self.id_skip = block_args.id_skip + self.expand_ratio = block_args.expand_ratio + self.drop_connect_rate = drop_connect_rate + self.is_test = is_test + + if self.expand_ratio != 1: + self._ecn = ExpandConvNorm(input_channels, + block_args, + padding_type=padding_type, + name=name, + model_name=model_name, + cur_stage=cur_stage) + + self._dcn = DepthwiseConvNorm(input_channels * block_args.expand_ratio, + block_args, + padding_type=padding_type, + name=name, + model_name=model_name, + cur_stage=cur_stage) + + if self.has_se: + num_squeezed_channels = max(1, int(block_args.input_filters * block_args.se_ratio)) + self._se = SEBlock(input_channels * block_args.expand_ratio, + num_squeezed_channels, + oup, + padding_type=padding_type, + name=name, + model_name=model_name, + cur_stage=cur_stage) + + self._pcn = ProjectConvNorm(input_channels * block_args.expand_ratio, + block_args, + padding_type=padding_type, + name=name, + model_name=model_name, + cur_stage=cur_stage) + + def forward(self, inputs: paddle.Tensor): + x = inputs + if self.expand_ratio != 1: + x = self._ecn(x) + x = F.swish(x) + x = self._dcn(x) + x = F.swish(x) + if self.has_se: + x = self._se(x) + x = self._pcn(x) + if self.id_skip and \ + self.block_args.stride == 1 and \ + self.block_args.input_filters == self.block_args.output_filters: + if self.drop_connect_rate: + x = _drop_connect(x, self.drop_connect_rate, self.is_test) + x = paddle.elementwise_add(x, inputs) + return x + + +class ConvStemNorm(nn.Layer): + """Basic conv stem norm block for extracting features.""" + def __init__(self, + input_channels: int, + padding_type: str, + _global_params: dict, + name: str = None, + model_name: str = None, + cur_stage: str = None): + super(ConvStemNorm, self).__init__() + + output_channels = round_filters(32, _global_params) + self._conv = ConvBNLayer(input_channels, + filter_size=3, + output_channels=output_channels, + stride=2, + bn_act=None, + padding_type=padding_type, + name="", + conv_name="_conv_stem", + bn_name="_bn0", + model_name=model_name, + cur_stage=cur_stage) + + def forward(self, inputs: paddle.Tensor): + return self._conv(inputs) + + +class ExtractFeatures(nn.Layer): + """Extract features.""" + def __init__(self, + input_channels: int, + _block_args: dict, + _global_params: dict, + padding_type: str, + use_se: bool, + is_test: bool, + model_name: str = None): + super(ExtractFeatures, self).__init__() + + self._global_params = _global_params + + self._conv_stem = ConvStemNorm(input_channels, + padding_type=padding_type, + _global_params=_global_params, + model_name=model_name, + cur_stage=0) + + self.block_args_copy = copy.deepcopy(_block_args) + idx = 0 + block_size = 0 + for block_arg in self.block_args_copy: + block_arg = block_arg._replace(input_filters=round_filters(block_arg.input_filters, _global_params), + output_filters=round_filters(block_arg.output_filters, _global_params), + num_repeat=round_repeats(block_arg.num_repeat, _global_params)) + block_size += 1 + for _ in range(block_arg.num_repeat - 1): + block_size += 1 + + self.conv_seq = [] + cur_stage = 1 + for block_args in _block_args: + block_args = block_args._replace(input_filters=round_filters(block_args.input_filters, _global_params), + output_filters=round_filters(block_args.output_filters, _global_params), + num_repeat=round_repeats(block_args.num_repeat, _global_params)) + + drop_connect_rate = self._global_params.drop_connect_rate if not is_test else 0 + if drop_connect_rate: + drop_connect_rate *= float(idx) / block_size + + _mc_block = self.add_sublayer( + "_blocks." + str(idx) + ".", + MbConvBlock(block_args.input_filters, + block_args=block_args, + padding_type=padding_type, + use_se=use_se, + name="_blocks." + str(idx) + ".", + drop_connect_rate=drop_connect_rate, + model_name=model_name, + cur_stage=cur_stage)) + self.conv_seq.append(_mc_block) + idx += 1 + if block_args.num_repeat > 1: + block_args = block_args._replace(input_filters=block_args.output_filters, stride=1) + for _ in range(block_args.num_repeat - 1): + drop_connect_rate = self._global_params.drop_connect_rate if not is_test else 0 + if drop_connect_rate: + drop_connect_rate *= float(idx) / block_size + _mc_block = self.add_sublayer( + "block." + str(idx) + ".", + MbConvBlock(block_args.input_filters, + block_args, + padding_type=padding_type, + use_se=use_se, + name="_blocks." + str(idx) + ".", + drop_connect_rate=drop_connect_rate, + model_name=model_name, + cur_stage=cur_stage)) + self.conv_seq.append(_mc_block) + idx += 1 + cur_stage += 1 + + def forward(self, inputs: paddle.Tensor): + x = self._conv_stem(inputs) + x = F.swish(x) + for _mc_block in self.conv_seq: + x = _mc_block(x) + return x + + +@moduleinfo(name="efficientnetb3_imagenet", + type="cv/classification", + author="paddlepaddle", + author_email="", + summary="efficientnetb3_imagenet is a classification model, " + "this module is trained with Imagenet dataset.", + version="1.1.0", + meta=ImageClassifierModule) +class EfficientNet_B3(nn.Layer): + def __init__(self, + is_test: bool = False, + padding_type: str = "SAME", + override_params: dict = None, + use_se: bool = True, + class_dim: int = 1000, + load_checkpoint: str = None): + super(EfficientNet_B3, self).__init__() + + model_name = 'efficientnet-b3' + self.name = "b3" + self._block_args, self._global_params = get_model_params(model_name, override_params) + self.padding_type = padding_type + self.use_se = use_se + self.is_test = is_test + + self._ef = ExtractFeatures(3, + self._block_args, + self._global_params, + self.padding_type, + self.use_se, + self.is_test, + model_name=self.name) + + output_channels = round_filters(1280, self._global_params) + oup = 384 + + self._conv = ConvBNLayer(oup, + 1, + output_channels, + bn_act="swish", + padding_type=self.padding_type, + name="", + conv_name="_conv_head", + bn_name="_bn1", + model_name=self.name, + cur_stage=7) + self._pool = AdaptiveAvgPool2d(1) + + if self._global_params.dropout_rate: + self._drop = Dropout(p=self._global_params.dropout_rate, mode="upscale_in_train") + + param_attr, bias_attr = init_fc_layer("_fc") + self._fc = Linear(output_channels, class_dim, weight_attr=param_attr, bias_attr=bias_attr) + + if load_checkpoint is not None: + model_dict = paddle.load(load_checkpoint)[0] + self.set_dict(model_dict) + print("load custom checkpoint success") + + else: + checkpoint = os.path.join(self.directory, 'efficientnet_b3_imagenet.pdparams') + if not os.path.exists(checkpoint): + os.system( + 'wget https://paddlehub.bj.bcebos.com/dygraph/image_classification/efficientnet_b3_imagenet.pdparams -O ' + + checkpoint) + model_dict = paddle.load(checkpoint)[0] + self.set_dict(model_dict) + print("load pretrained checkpoint success") + + def forward(self, inputs: paddle.Tensor): + x = self._ef(inputs) + x = self._conv(x) + x = self._pool(x) + if self._global_params.dropout_rate: + x = self._drop(x) + x = paddle.squeeze(x, axis=[2, 3]) + x = self._fc(x) + return x diff --git a/hub_module/modules/image/classification/efficientnetb3_imagenet/processor.py b/hub_module/modules/image/classification/efficientnetb3_imagenet/processor.py deleted file mode 100644 index 12ccc9b2..00000000 --- a/hub_module/modules/image/classification/efficientnetb3_imagenet/processor.py +++ /dev/null @@ -1,69 +0,0 @@ -# -*- coding:utf-8 -*- -# copyright (c) 2020 PaddlePaddle Authors. All Rights Reserve. -# -# Licensed under the Apache License, Version 2.0 (the "License"); -# you may not use this file except in compliance with the License. -# You may obtain a copy of the License at -# -# http://www.apache.org/licenses/LICENSE-2.0 -# -# Unless required by applicable law or agreed to in writing, software -# distributed under the License is distributed on an "AS IS" BASIS, -# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. -# See the License for the specific language governing permissions and -# limitations under the License. - -from __future__ import absolute_import -from __future__ import division -from __future__ import print_function - -import base64 -import cv2 -import os - -import numpy as np - - -def base64_to_cv2(b64str): - data = base64.b64decode(b64str.encode('utf8')) - data = np.fromstring(data, np.uint8) - data = cv2.imdecode(data, cv2.IMREAD_COLOR) - return data - - -def softmax(x): - orig_shape = x.shape - if len(x.shape) > 1: - tmp = np.max(x, axis=1) - x -= tmp.reshape((x.shape[0], 1)) - x = np.exp(x) - tmp = np.sum(x, axis=1) - x /= tmp.reshape((x.shape[0], 1)) - else: - tmp = np.max(x) - x -= tmp - x = np.exp(x) - tmp = np.sum(x) - x /= tmp - return x - - -def postprocess(data_out, label_list, top_k): - """ - Postprocess output of network, one image at a time. - - Args: - data_out (numpy.ndarray): output data of network. - label_list (list): list of label. - top_k (int): Return top k results. - """ - output = [] - for result in data_out: - result_i = softmax(result) - output_i = {} - indexs = np.argsort(result_i)[::-1][0:top_k] - for index in indexs: - label = label_list[index].split(',')[0] - output_i[label] = float(result_i[index]) - output.append(output_i) - return output diff --git a/hub_module/modules/image/classification/efficientnetb4_imagenet/README.md b/hub_module/modules/image/classification/efficientnetb4_imagenet/README.md deleted file mode 100644 index aded214a..00000000 --- a/hub_module/modules/image/classification/efficientnetb4_imagenet/README.md +++ /dev/null @@ -1,149 +0,0 @@ -## 命令行预测 - -``` -hub run efficientnetb4_imagenet --input_path "/PATH/TO/IMAGE" -``` - -## API - -```python -def get_expected_image_width() -``` - -返回预处理的图片宽度，也就是224。 - -```python -def get_expected_image_height() -``` - -返回预处理的图片高度，也就是224。 - -```python -def get_pretrained_images_mean() -``` - -返回预处理的图片均值，也就是 \[0.485, 0.456, 0.406\]。 - -```python -def get_pretrained_images_std() -``` - -返回预处理的图片标准差，也就是 \[0.229, 0.224, 0.225\]。 - - -```python -def context(trainable=True, pretrained=True) -``` - -**参数** - -* trainable (bool): 计算图的参数是否为可训练的； -* pretrained (bool): 是否加载默认的预训练模型。 - -**返回** - -* inputs (dict): 计算图的输入，key 为 'image', value 为图片的张量； -* outputs (dict): 计算图的输出，key 为 'classification' 和 'feature_map'，其相应的值为： - * classification (paddle.fluid.framework.Variable): 分类结果，也就是全连接层的输出； - * feature\_map (paddle.fluid.framework.Variable): 特征匹配，全连接层前面的那个张量。 -* context\_prog(fluid.Program): 计算图，用于迁移学习。 - -```python -def classify(images=None, - paths=None, - batch_size=1, - use_gpu=False, - top_k=1): -``` - -**参数** - -* images (list\[numpy.ndarray\]): 图片数据，每一个图片数据的shape 均为 \[H, W, C\]，颜色空间为 BGR； -* paths (list\[str\]): 图片的路径； -* batch\_size (int): batch 的大小； -* use\_gpu (bool): 是否使用 GPU 来预测； -* top\_k (int): 返回预测结果的前 k 个。 - -**返回** - -res (list\[dict\]): 分类结果，列表的每一个元素均为字典，其中 key 为识别动物的类别，value为置信度。 - -```python -def save_inference_model(dirname, - model_filename=None, - params_filename=None, - combined=True) -``` - -将模型保存到指定路径。 - -**参数** - -* dirname: 存在模型的目录名称 -* model\_filename: 模型文件名称，默认为\_\_model\_\_ -* params\_filename: 参数文件名称，默认为\_\_params\_\_(仅当`combined`为True时生效) -* combined: 是否将参数保存到统一的一个文件中 - -## 代码示例 - -```python -import paddlehub as hub -import cv2 - -classifier = hub.Module(name="efficientnetb4_imagenet") - -result = classifier.classify(images=[cv2.imread('/PATH/TO/IMAGE')]) -# or -# result = classifier.classify(paths=['/PATH/TO/IMAGE']) -``` - -## 服务部署 - -PaddleHub Serving可以部署一个在线图像识别服务。 - -## 第一步：启动PaddleHub Serving - -运行启动命令： -```shell -$ hub serving start -m efficientnetb4_imagenet -``` - -这样就完成了一个在线图像识别服务化API的部署，默认端口号为8866。 - -**NOTE:** 如使用GPU预测，则需要在启动服务之前，请设置CUDA\_VISIBLE\_DEVICES环境变量，否则不用设置。 - -## 第二步：发送预测请求 - -配置好服务端，以下数行代码即可实现发送预测请求，获取预测结果 - -```python -import requests -import json -import cv2 -import base64 - - -def cv2_to_base64(image): - data = cv2.imencode('.jpg', image)[1] - return base64.b64encode(data.tostring()).decode('utf8') - - -# 发送HTTP请求 -data = {'images':[cv2_to_base64(cv2.imread("/PATH/TO/IMAGE"))]} -headers = {"Content-type": "application/json"} -url = "http://127.0.0.1:8866/predict/efficientnetb4_imagenet" -r = requests.post(url=url, headers=headers, data=json.dumps(data)) - -# 打印预测结果 -print(r.json()["results"]) -``` - -### 查看代码 - -https://github.com/PaddlePaddle/PaddleClas - -### 依赖 - -paddlepaddle >= 1.6.2 - -paddlehub >= 1.6.0 diff --git a/hub_module/modules/image/classification/efficientnetb4_imagenet/__init__.py b/hub_module/modules/image/classification/efficientnetb4_imagenet/__init__.py deleted file mode 100644 index 19a990da..00000000 --- a/hub_module/modules/image/classification/efficientnetb4_imagenet/__init__.py +++ /dev/null @@ -1 +0,0 @@ -# -*- coding:utf-8 -*- diff --git a/hub_module/modules/image/classification/efficientnetb4_imagenet/data_feed.py b/hub_module/modules/image/classification/efficientnetb4_imagenet/data_feed.py deleted file mode 100644 index 11181458..00000000 --- a/hub_module/modules/image/classification/efficientnetb4_imagenet/data_feed.py +++ /dev/null @@ -1,98 +0,0 @@ -# -*- coding:utf-8 -*- -# copyright (c) 2020 PaddlePaddle Authors. All Rights Reserve. -# -# Licensed under the Apache License, Version 2.0 (the "License"); -# you may not use this file except in compliance with the License. -# You may obtain a copy of the License at -# -# http://www.apache.org/licenses/LICENSE-2.0 -# -# Unless required by applicable law or agreed to in writing, software -# distributed under the License is distributed on an "AS IS" BASIS, -# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. -# See the License for the specific language governing permissions and -# limitations under the License. - -import os -import time -from collections import OrderedDict - -import cv2 -import numpy as np -from PIL import Image - -__all__ = ['reader'] - -DATA_DIM = 224 -img_mean = np.array([0.485, 0.456, 0.406]).reshape((3, 1, 1)) -img_std = np.array([0.229, 0.224, 0.225]).reshape((3, 1, 1)) - - -def resize_short(img, target_size): - percent = float(target_size) / min(img.size[0], img.size[1]) - resized_width = int(round(img.size[0] * percent)) - resized_height = int(round(img.size[1] * percent)) - img = img.resize((resized_width, resized_height), Image.LANCZOS) - return img - - -def crop_image(img, target_size, center): - width, height = img.size - size = target_size - if center == True: - w_start = (width - size) / 2 - h_start = (height - size) / 2 - else: - w_start = np.random.randint(0, width - size + 1) - h_start = np.random.randint(0, height - size + 1) - w_end = w_start + size - h_end = h_start + size - img = img.crop((w_start, h_start, w_end, h_end)) - return img - - -def process_image(img): - img = resize_short(img, target_size=256) - img = crop_image(img, target_size=DATA_DIM, center=True) - if img.mode != 'RGB': - img = img.convert('RGB') - img = np.array(img).astype('float32').transpose((2, 0, 1)) / 255 - img -= img_mean - img /= img_std - return img - - -def reader(images=None, paths=None): - """ - Preprocess to yield image. - - Args: - images (list[numpy.ndarray]): images data, shape of each is [H, W, C]. - paths (list[str]): paths to images. - - Yield: - each (collections.OrderedDict): info of original image, preprocessed image. - """ - component = list() - if paths: - for im_path in paths: - each = OrderedDict() - assert os.path.isfile( - im_path), "The {} isn't a valid file path.".format(im_path) - each['org_im_path'] = im_path - each['org_im'] = Image.open(im_path) - each['org_im_width'], each['org_im_height'] = each['org_im'].size - component.append(each) - if images is not None: - assert type(images), "images is a list." - for im in images: - each = OrderedDict() - each['org_im'] = Image.fromarray(im[:, :, ::-1]) - each['org_im_path'] = 'ndarray_time={}'.format( - round(time.time(), 6) * 1e6) - each['org_im_width'], each['org_im_height'] = each['org_im'].size - component.append(each) - - for element in component: - element['image'] = process_image(element['org_im']) - yield element diff --git a/hub_module/modules/image/classification/efficientnetb4_imagenet/efficientnet.py b/hub_module/modules/image/classification/efficientnetb4_imagenet/efficientnet.py deleted file mode 100644 index f3684529..00000000 --- a/hub_module/modules/image/classification/efficientnetb4_imagenet/efficientnet.py +++ /dev/null @@ -1,623 +0,0 @@ -# -*- coding:utf-8 -*- -# copyright (c) 2020 PaddlePaddle Authors. All Rights Reserve. -# -# Licensed under the Apache License, Version 2.0 (the "License"); -# you may not use this file except in compliance with the License. -# You may obtain a copy of the License at -# -# http://www.apache.org/licenses/LICENSE-2.0 -# -# Unless required by applicable law or agreed to in writing, software -# distributed under the License is distributed on an "AS IS" BASIS, -# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. -# See the License for the specific language governing permissions and -# limitations under the License. - -from __future__ import absolute_import -from __future__ import division -from __future__ import print_function - -import collections -import re -import math -import copy - -import paddle.fluid as fluid -from efficientnetb4_imagenet.layers import conv2d, init_batch_norm_layer, init_fc_layer - -__all__ = [ - 'EfficientNet', 'EfficientNetB0_small', 'EfficientNetB0', 'EfficientNetB1', - 'EfficientNetB2', 'EfficientNetB3', 'EfficientNetB4', 'EfficientNetB5', - 'EfficientNetB6', 'EfficientNetB7' -] - -GlobalParams = collections.namedtuple('GlobalParams', [ - 'batch_norm_momentum', - 'batch_norm_epsilon', - 'dropout_rate', - 'num_classes', - 'width_coefficient', - 'depth_coefficient', - 'depth_divisor', - 'min_depth', - 'drop_connect_rate', -]) - -BlockArgs = collections.namedtuple('BlockArgs', [ - 'kernel_size', 'num_repeat', 'input_filters', 'output_filters', - 'expand_ratio', 'id_skip', 'stride', 'se_ratio' -]) - -GlobalParams.__new__.__defaults__ = (None, ) * len(GlobalParams._fields) -BlockArgs.__new__.__defaults__ = (None, ) * len(BlockArgs._fields) - - -def efficientnet_params(model_name): - """ Map EfficientNet model name to parameter coefficients. """ - params_dict = { - # Coefficients: width,depth,resolution,dropout - 'efficientnet-b0': (1.0, 1.0, 224, 0.2), - 'efficientnet-b1': (1.0, 1.1, 240, 0.2), - 'efficientnet-b2': (1.1, 1.2, 260, 0.3), - 'efficientnet-b3': (1.2, 1.4, 300, 0.3), - 'efficientnet-b4': (1.4, 1.8, 380, 0.4), - 'efficientnet-b5': (1.6, 2.2, 456, 0.4), - 'efficientnet-b6': (1.8, 2.6, 528, 0.5), - 'efficientnet-b7': (2.0, 3.1, 600, 0.5), - } - return params_dict[model_name] - - -def efficientnet(width_coefficient=None, - depth_coefficient=None, - dropout_rate=0.2, - drop_connect_rate=0.2): - """ Get block arguments according to parameter and coefficients. """ - blocks_args = [ - 'r1_k3_s11_e1_i32_o16_se0.25', - 'r2_k3_s22_e6_i16_o24_se0.25', - 'r2_k5_s22_e6_i24_o40_se0.25', - 'r3_k3_s22_e6_i40_o80_se0.25', - 'r3_k5_s11_e6_i80_o112_se0.25', - 'r4_k5_s22_e6_i112_o192_se0.25', - 'r1_k3_s11_e6_i192_o320_se0.25', - ] - blocks_args = BlockDecoder.decode(blocks_args) - - global_params = GlobalParams( - batch_norm_momentum=0.99, - batch_norm_epsilon=1e-3, - dropout_rate=dropout_rate, - drop_connect_rate=drop_connect_rate, - num_classes=1000, - width_coefficient=width_coefficient, - depth_coefficient=depth_coefficient, - depth_divisor=8, - min_depth=None) - - return blocks_args, global_params - - -def get_model_params(model_name, override_params): - """ Get the block args and global params for a given model """ - if model_name.startswith('efficientnet'): - w, d, _, p = efficientnet_params(model_name) - blocks_args, global_params = efficientnet( - width_coefficient=w, depth_coefficient=d, dropout_rate=p) - else: - raise NotImplementedError( - 'model name is not pre-defined: %s' % model_name) - if override_params: - global_params = global_params._replace(**override_params) - return blocks_args, global_params - - -def round_filters(filters, global_params): - """ Calculate and round number of filters based on depth multiplier. """ - multiplier = global_params.width_coefficient - if not multiplier: - return filters - divisor = global_params.depth_divisor - min_depth = global_params.min_depth - filters *= multiplier - min_depth = min_depth or divisor - new_filters = max(min_depth, - int(filters + divisor / 2) // divisor * divisor) - if new_filters < 0.9 * filters: # prevent rounding by more than 10% - new_filters += divisor - return int(new_filters) - - -def round_repeats(repeats, global_params): - """ Round number of filters based on depth multiplier. """ - multiplier = global_params.depth_coefficient - if not multiplier: - return repeats - return int(math.ceil(multiplier * repeats)) - - -class EfficientNet(): - def __init__(self, - name='b0', - padding_type='SAME', - override_params=None, - is_test=False, - use_se=True): - valid_names = ['b' + str(i) for i in range(8)] - assert name in valid_names, 'efficient name should be in b0~b7' - model_name = 'efficientnet-' + name - self._blocks_args, self._global_params = get_model_params( - model_name, override_params) - self._bn_mom = self._global_params.batch_norm_momentum - self._bn_eps = self._global_params.batch_norm_epsilon - self.padding_type = padding_type - self.use_se = use_se - - def net(self, input, class_dim=1000, is_test=False): - - conv = self.extract_features(input, is_test=is_test) - - out_channels = round_filters(1280, self._global_params) - conv = self.conv_bn_layer( - conv, - num_filters=out_channels, - filter_size=1, - bn_act='swish', - bn_mom=self._bn_mom, - bn_eps=self._bn_eps, - padding_type=self.padding_type, - name='', - conv_name='_conv_head', - bn_name='_bn1') - - pool = fluid.layers.pool2d( - input=conv, pool_type='avg', global_pooling=True, use_cudnn=False) - - if not is_test and self._global_params.dropout_rate: - pool = fluid.layers.dropout( - pool, - self._global_params.dropout_rate, - dropout_implementation='upscale_in_train') - - param_attr, bias_attr = init_fc_layer(class_dim, '_fc') - out = fluid.layers.fc( - pool, - class_dim, - name='_fc', - param_attr=param_attr, - bias_attr=bias_attr) - return out, pool - - def _drop_connect(self, inputs, prob, is_test): - if is_test: - return inputs - keep_prob = 1.0 - prob - random_tensor = keep_prob + fluid.layers.uniform_random_batch_size_like( - inputs, [-1, 1, 1, 1], min=0., max=1.) - binary_tensor = fluid.layers.floor(random_tensor) - output = inputs / keep_prob * binary_tensor - return output - - def _expand_conv_norm(self, inputs, block_args, is_test, name=None): - # Expansion phase - oup = block_args.input_filters * block_args.expand_ratio # number of output channels - - if block_args.expand_ratio != 1: - conv = self.conv_bn_layer( - inputs, - num_filters=oup, - filter_size=1, - bn_act=None, - bn_mom=self._bn_mom, - bn_eps=self._bn_eps, - padding_type=self.padding_type, - name=name, - conv_name=name + '_expand_conv', - bn_name='_bn0') - - return conv - - def _depthwise_conv_norm(self, inputs, block_args, is_test, name=None): - k = block_args.kernel_size - s = block_args.stride - if isinstance(s, list) or isinstance(s, tuple): - s = s[0] - oup = block_args.input_filters * block_args.expand_ratio # number of output channels - - conv = self.conv_bn_layer( - inputs, - num_filters=oup, - filter_size=k, - stride=s, - num_groups=oup, - bn_act=None, - padding_type=self.padding_type, - bn_mom=self._bn_mom, - bn_eps=self._bn_eps, - name=name, - use_cudnn=False, - conv_name=name + '_depthwise_conv', - bn_name='_bn1') - - return conv - - def _project_conv_norm(self, inputs, block_args, is_test, name=None): - final_oup = block_args.output_filters - conv = self.conv_bn_layer( - inputs, - num_filters=final_oup, - filter_size=1, - bn_act=None, - padding_type=self.padding_type, - bn_mom=self._bn_mom, - bn_eps=self._bn_eps, - name=name, - conv_name=name + '_project_conv', - bn_name='_bn2') - return conv - - def conv_bn_layer(self, - input, - filter_size, - num_filters, - stride=1, - num_groups=1, - padding_type="SAME", - conv_act=None, - bn_act='swish', - use_cudnn=True, - use_bn=True, - bn_mom=0.9, - bn_eps=1e-05, - use_bias=False, - name=None, - conv_name=None, - bn_name=None): - conv = conv2d( - input=input, - num_filters=num_filters, - filter_size=filter_size, - stride=stride, - groups=num_groups, - act=conv_act, - padding_type=padding_type, - use_cudnn=use_cudnn, - name=conv_name, - use_bias=use_bias) - - if use_bn == False: - return conv - else: - bn_name = name + bn_name - param_attr, bias_attr = init_batch_norm_layer(bn_name) - return fluid.layers.batch_norm( - input=conv, - act=bn_act, - momentum=bn_mom, - epsilon=bn_eps, - name=bn_name, - moving_mean_name=bn_name + '_mean', - moving_variance_name=bn_name + '_variance', - param_attr=param_attr, - bias_attr=bias_attr) - - def _conv_stem_norm(self, inputs, is_test): - out_channels = round_filters(32, self._global_params) - bn = self.conv_bn_layer( - inputs, - num_filters=out_channels, - filter_size=3, - stride=2, - bn_act=None, - bn_mom=self._bn_mom, - padding_type=self.padding_type, - bn_eps=self._bn_eps, - name='', - conv_name='_conv_stem', - bn_name='_bn0') - - return bn - - def mb_conv_block(self, - inputs, - block_args, - is_test=False, - drop_connect_rate=None, - name=None): - # Expansion and Depthwise Convolution - oup = block_args.input_filters * block_args.expand_ratio # number of output channels - has_se = self.use_se and (block_args.se_ratio is - not None) and (0 < block_args.se_ratio <= 1) - id_skip = block_args.id_skip # skip connection and drop connect - conv = inputs - if block_args.expand_ratio != 1: - conv = fluid.layers.swish( - self._expand_conv_norm(conv, block_args, is_test, name)) - - conv = fluid.layers.swish( - self._depthwise_conv_norm(conv, block_args, is_test, name)) - - # Squeeze and Excitation - if has_se: - num_squeezed_channels = max( - 1, int(block_args.input_filters * block_args.se_ratio)) - conv = self.se_block(conv, num_squeezed_channels, oup, name) - - conv = self._project_conv_norm(conv, block_args, is_test, name) - - # Skip connection and drop connect - input_filters, output_filters = block_args.input_filters, block_args.output_filters - if id_skip and block_args.stride == 1 and input_filters == output_filters: - if drop_connect_rate: - conv = self._drop_connect(conv, drop_connect_rate, is_test) - conv = fluid.layers.elementwise_add(conv, inputs) - - return conv - - def se_block(self, inputs, num_squeezed_channels, oup, name): - x_squeezed = fluid.layers.pool2d( - input=inputs, pool_type='avg', global_pooling=True, use_cudnn=False) - x_squeezed = conv2d( - x_squeezed, - num_filters=num_squeezed_channels, - filter_size=1, - use_bias=True, - padding_type=self.padding_type, - act='swish', - name=name + '_se_reduce') - x_squeezed = conv2d( - x_squeezed, - num_filters=oup, - filter_size=1, - use_bias=True, - padding_type=self.padding_type, - name=name + '_se_expand') - se_out = inputs * fluid.layers.sigmoid(x_squeezed) - return se_out - - def extract_features(self, inputs, is_test): - """ Returns output of the final convolution layer """ - - conv = fluid.layers.swish(self._conv_stem_norm(inputs, is_test=is_test)) - - block_args_copy = copy.deepcopy(self._blocks_args) - idx = 0 - block_size = 0 - for block_arg in block_args_copy: - block_arg = block_arg._replace( - input_filters=round_filters(block_arg.input_filters, - self._global_params), - output_filters=round_filters(block_arg.output_filters, - self._global_params), - num_repeat=round_repeats(block_arg.num_repeat, - self._global_params)) - block_size += 1 - for _ in range(block_arg.num_repeat - 1): - block_size += 1 - - for block_args in self._blocks_args: - - # Update block input and output filters based on depth multiplier. - block_args = block_args._replace( - input_filters=round_filters(block_args.input_filters, - self._global_params), - output_filters=round_filters(block_args.output_filters, - self._global_params), - num_repeat=round_repeats(block_args.num_repeat, - self._global_params)) - - # The first block needs to take care of stride and filter size increase. - drop_connect_rate = self._global_params.drop_connect_rate - if drop_connect_rate: - drop_connect_rate *= float(idx) / block_size - conv = self.mb_conv_block(conv, block_args, is_test, - drop_connect_rate, - '_blocks.' + str(idx) + '.') - - idx += 1 - if block_args.num_repeat > 1: - block_args = block_args._replace( - input_filters=block_args.output_filters, stride=1) - for _ in range(block_args.num_repeat - 1): - drop_connect_rate = self._global_params.drop_connect_rate - if drop_connect_rate: - drop_connect_rate *= float(idx) / block_size - conv = self.mb_conv_block(conv, block_args, is_test, - drop_connect_rate, - '_blocks.' + str(idx) + '.') - idx += 1 - - return conv - - def shortcut(self, input, data_residual): - return fluid.layers.elementwise_add(input, data_residual) - - -class BlockDecoder(object): - """ Block Decoder for readability, straight from the official TensorFlow repository """ - - @staticmethod - def _decode_block_string(block_string): - """ Gets a block through a string notation of arguments. """ - assert isinstance(block_string, str) - - ops = block_string.split('_') - options = {} - for op in ops: - splits = re.split(r'(\d.*)', op) - if len(splits) >= 2: - key, value = splits[:2] - options[key] = value - - # Check stride - assert (('s' in options and len(options['s']) == 1) or - (len(options['s']) == 2 and options['s'][0] == options['s'][1])) - - return BlockArgs( - kernel_size=int(options['k']), - num_repeat=int(options['r']), - input_filters=int(options['i']), - output_filters=int(options['o']), - expand_ratio=int(options['e']), - id_skip=('noskip' not in block_string), - se_ratio=float(options['se']) if 'se' in options else None, - stride=[int(options['s'][0])]) - - @staticmethod - def _encode_block_string(block): - """Encodes a block to a string.""" - args = [ - 'r%d' % block.num_repeat, - 'k%d' % block.kernel_size, - 's%d%d' % (block.strides[0], block.strides[1]), - 'e%s' % block.expand_ratio, - 'i%d' % block.input_filters, - 'o%d' % block.output_filters - ] - if 0 < block.se_ratio <= 1: - args.append('se%s' % block.se_ratio) - if block.id_skip is False: - args.append('noskip') - return '_'.join(args) - - @staticmethod - def decode(string_list): - """ - Decodes a list of string notations to specify blocks inside the network. - :param string_list: a list of strings, each string is a notation of block - :return: a list of BlockArgs namedtuples of block args - """ - assert isinstance(string_list, list) - blocks_args = [] - for block_string in string_list: - blocks_args.append(BlockDecoder._decode_block_string(block_string)) - return blocks_args - - @staticmethod - def encode(blocks_args): - """ - Encodes a list of BlockArgs to a list of strings. - :param blocks_args: a list of BlockArgs namedtuples of block args - :return: a list of strings, each string is a notation of block - """ - block_strings = [] - for block in blocks_args: - block_strings.append(BlockDecoder._encode_block_string(block)) - return block_strings - - -def EfficientNetB0_small(is_test=False, - padding_type='SAME', - override_params=None, - use_se=False): - model = EfficientNet( - name='b0', - is_test=is_test, - padding_type=padding_type, - override_params=override_params, - use_se=use_se) - return model - - -def EfficientNetB0(is_test=False, - padding_type='SAME', - override_params=None, - use_se=True): - model = EfficientNet( - name='b0', - is_test=is_test, - padding_type=padding_type, - override_params=override_params, - use_se=use_se) - return model - - -def EfficientNetB1(is_test=False, - padding_type='SAME', - override_params=None, - use_se=True): - model = EfficientNet( - name='b1', - is_test=is_test, - padding_type=padding_type, - override_params=override_params, - use_se=use_se) - return model - - -def EfficientNetB2(is_test=False, - padding_type='SAME', - override_params=None, - use_se=True): - model = EfficientNet( - name='b2', - is_test=is_test, - padding_type=padding_type, - override_params=override_params, - use_se=use_se) - return model - - -def EfficientNetB3(is_test=False, - padding_type='SAME', - override_params=None, - use_se=True): - model = EfficientNet( - name='b3', - is_test=is_test, - padding_type=padding_type, - override_params=override_params, - use_se=use_se) - return model - - -def EfficientNetB4(is_test=False, - padding_type='SAME', - override_params=None, - use_se=True): - model = EfficientNet( - name='b4', - is_test=is_test, - padding_type=padding_type, - override_params=override_params, - use_se=use_se) - return model - - -def EfficientNetB5(is_test=False, - padding_type='SAME', - override_params=None, - use_se=True): - model = EfficientNet( - name='b5', - is_test=is_test, - padding_type=padding_type, - override_params=override_params, - use_se=use_se) - return model - - -def EfficientNetB6(is_test=False, - padding_type='SAME', - override_params=None, - use_se=True): - model = EfficientNet( - name='b6', - is_test=is_test, - padding_type=padding_type, - override_params=override_params, - use_se=use_se) - return model - - -def EfficientNetB7(is_test=False, - padding_type='SAME', - override_params=None, - use_se=True): - model = EfficientNet( - name='b7', - is_test=is_test, - padding_type=padding_type, - override_params=override_params, - use_se=use_se) - return model diff --git a/hub_module/modules/image/classification/efficientnetb4_imagenet/label_list.txt b/hub_module/modules/image/classification/efficientnetb4_imagenet/label_list.txt deleted file mode 100644 index a509c007..00000000 --- a/hub_module/modules/image/classification/efficientnetb4_imagenet/label_list.txt +++ /dev/null @@ -1,1000 +0,0 @@ -tench, Tinca tinca -goldfish, Carassius auratus -great white shark, white shark, man-eater, man-eating shark, Carcharodon carcharias -tiger shark, Galeocerdo cuvieri -hammerhead, hammerhead shark -electric ray, crampfish, numbfish, torpedo -stingray -cock -hen -ostrich, Struthio camelus -brambling, Fringilla montifringilla -goldfinch, Carduelis carduelis -house finch, linnet, Carpodacus mexicanus -junco, snowbird -indigo bunting, indigo finch, indigo bird, Passerina cyanea -robin, American robin, Turdus migratorius -bulbul -jay -magpie -chickadee -water ouzel, dipper -kite -bald eagle, American eagle, Haliaeetus leucocephalus -vulture -great grey owl, great gray owl, Strix nebulosa -European fire salamander, Salamandra salamandra -common newt, Triturus vulgaris -eft -spotted salamander, Ambystoma maculatum -axolotl, mud puppy, Ambystoma mexicanum -bullfrog, Rana catesbeiana -tree frog, tree-frog -tailed frog, bell toad, ribbed toad, tailed toad, Ascaphus trui -loggerhead, loggerhead turtle, Caretta caretta -leatherback turtle, leatherback, leathery turtle, Dermochelys coriacea -mud turtle -terrapin -box turtle, box tortoise -banded gecko -common iguana, iguana, Iguana iguana -American chameleon, anole, Anolis carolinensis -whiptail, whiptail lizard -agama -frilled lizard, Chlamydosaurus kingi -alligator lizard -Gila monster, Heloderma suspectum -green lizard, Lacerta viridis -African chameleon, Chamaeleo chamaeleon -Komodo dragon, Komodo lizard, dragon lizard, giant lizard, Varanus komodoensis -African crocodile, Nile crocodile, Crocodylus niloticus -American alligator, Alligator mississipiensis -triceratops -thunder snake, worm snake, Carphophis amoenus -ringneck snake, ring-necked snake, ring snake -hognose snake, puff adder, sand viper -green snake, grass snake -king snake, kingsnake -garter snake, grass snake -water snake -vine snake -night snake, Hypsiglena torquata -boa constrictor, Constrictor constrictor -rock python, rock snake, Python sebae -Indian cobra, Naja naja -green mamba -sea snake -horned viper, cerastes, sand viper, horned asp, Cerastes cornutus -diamondback, diamondback rattlesnake, Crotalus adamanteus -sidewinder, horned rattlesnake, Crotalus cerastes -trilobite -harvestman, daddy longlegs, Phalangium opilio -scorpion -black and gold garden spider, Argiope aurantia -barn spider, Araneus cavaticus -garden spider, Aranea diademata -black widow, Latrodectus mactans -tarantula -wolf spider, hunting spider -tick -centipede -black grouse -ptarmigan -ruffed grouse, partridge, Bonasa umbellus -prairie chicken, prairie grouse, prairie fowl -peacock -quail -partridge -African grey, African gray, Psittacus erithacus -macaw -sulphur-crested cockatoo, Kakatoe galerita, Cacatua galerita -lorikeet -coucal -bee eater -hornbill -hummingbird -jacamar -toucan -drake -red-breasted merganser, Mergus serrator -goose -black swan, Cygnus atratus -tusker -echidna, spiny anteater, anteater -platypus, duckbill, duckbilled platypus, duck-billed platypus, Ornithorhynchus anatinus -wallaby, brush kangaroo -koala, koala bear, kangaroo bear, native bear, Phascolarctos cinereus -wombat -jellyfish -sea anemone, anemone -brain coral -flatworm, platyhelminth -nematode, nematode worm, roundworm -conch -snail -slug -sea slug, nudibranch -chiton, coat-of-mail shell, sea cradle, polyplacophore -chambered nautilus, pearly nautilus, nautilus -Dungeness crab, Cancer magister -rock crab, Cancer irroratus -fiddler crab -king crab, Alaska crab, Alaskan king crab, Alaska king crab, Paralithodes camtschatica -American lobster, Northern lobster, Maine lobster, Homarus americanus -spiny lobster, langouste, rock lobster, crawfish, crayfish, sea crawfish -crayfish, crawfish, crawdad, crawdaddy -hermit crab -isopod -white stork, Ciconia ciconia -black stork, Ciconia nigra -spoonbill -flamingo -little blue heron, Egretta caerulea -American egret, great white heron, Egretta albus -bittern -crane -limpkin, Aramus pictus -European gallinule, Porphyrio porphyrio -American coot, marsh hen, mud hen, water hen, Fulica americana -bustard -ruddy turnstone, Arenaria interpres -red-backed sandpiper, dunlin, Erolia alpina -redshank, Tringa totanus -dowitcher -oystercatcher, oyster catcher -pelican -king penguin, Aptenodytes patagonica -albatross, mollymawk -grey whale, gray whale, devilfish, Eschrichtius gibbosus, Eschrichtius robustus -killer whale, killer, orca, grampus, sea wolf, Orcinus orca -dugong, Dugong dugon -sea lion -Chihuahua -Japanese spaniel -Maltese dog, Maltese terrier, Maltese -Pekinese, Pekingese, Peke -Shih-Tzu -Blenheim spaniel -papillon -toy terrier -Rhodesian ridgeback -Afghan hound, Afghan -basset, basset hound -beagle -bloodhound, sleuthhound -bluetick -black-and-tan coonhound -Walker hound, Walker foxhound -English foxhound -redbone -borzoi, Russian wolfhound -Irish wolfhound -Italian greyhound -whippet -Ibizan hound, Ibizan Podenco -Norwegian elkhound, elkhound -otterhound, otter hound -Saluki, gazelle hound -Scottish deerhound, deerhound -Weimaraner -Staffordshire bullterrier, Staffordshire bull terrier -American Staffordshire terrier, Staffordshire terrier, American pit bull terrier, pit bull terrier -Bedlington terrier -Border terrier -Kerry blue terrier -Irish terrier -Norfolk terrier -Norwich terrier -Yorkshire terrier -wire-haired fox terrier -Lakeland terrier -Sealyham terrier, Sealyham -Airedale, Airedale terrier -cairn, cairn terrier -Australian terrier -Dandie Dinmont, Dandie Dinmont terrier -Boston bull, Boston terrier -miniature schnauzer -giant schnauzer -standard schnauzer -Scotch terrier, Scottish terrier, Scottie -Tibetan terrier, chrysanthemum dog -silky terrier, Sydney silky -soft-coated wheaten terrier -West Highland white terrier -Lhasa, Lhasa apso -flat-coated retriever -curly-coated retriever -golden retriever -Labrador retriever -Chesapeake Bay retriever -German short-haired pointer -vizsla, Hungarian pointer -English setter -Irish setter, red setter -Gordon setter -Brittany spaniel -clumber, clumber spaniel -English springer, English springer spaniel -Welsh springer spaniel -cocker spaniel, English cocker spaniel, cocker -Sussex spaniel -Irish water spaniel -kuvasz -schipperke -groenendael -malinois -briard -kelpie -komondor -Old English sheepdog, bobtail -Shetland sheepdog, Shetland sheep dog, Shetland -collie -Border collie -Bouvier des Flandres, Bouviers des Flandres -Rottweiler -German shepherd, German shepherd dog, German police dog, alsatian -Doberman, Doberman pinscher -miniature pinscher -Greater Swiss Mountain dog -Bernese mountain dog -Appenzeller -EntleBucher -boxer -bull mastiff -Tibetan mastiff -French bulldog -Great Dane -Saint Bernard, St Bernard -Eskimo dog, husky -malamute, malemute, Alaskan malamute -Siberian husky -dalmatian, coach dog, carriage dog -affenpinscher, monkey pinscher, monkey dog -basenji -pug, pug-dog -Leonberg -Newfoundland, Newfoundland dog -Great Pyrenees -Samoyed, Samoyede -Pomeranian -chow, chow chow -keeshond -Brabancon griffon -Pembroke, Pembroke Welsh corgi -Cardigan, Cardigan Welsh corgi -toy poodle -miniature poodle -standard poodle -Mexican hairless -timber wolf, grey wolf, gray wolf, Canis lupus -white wolf, Arctic wolf, Canis lupus tundrarum -red wolf, maned wolf, Canis rufus, Canis niger -coyote, prairie wolf, brush wolf, Canis latrans -dingo, warrigal, warragal, Canis dingo -dhole, Cuon alpinus -African hunting dog, hyena dog, Cape hunting dog, Lycaon pictus -hyena, hyaena -red fox, Vulpes vulpes -kit fox, Vulpes macrotis -Arctic fox, white fox, Alopex lagopus -grey fox, gray fox, Urocyon cinereoargenteus -tabby, tabby cat -tiger cat -Persian cat -Siamese cat, Siamese -Egyptian cat -cougar, puma, catamount, mountain lion, painter, panther, Felis concolor -lynx, catamount -leopard, Panthera pardus -snow leopard, ounce, Panthera uncia -jaguar, panther, Panthera onca, Felis onca -lion, king of beasts, Panthera leo -tiger, Panthera tigris -cheetah, chetah, Acinonyx jubatus -brown bear, bruin, Ursus arctos -American black bear, black bear, Ursus americanus, Euarctos americanus -ice bear, polar bear, Ursus Maritimus, Thalarctos maritimus -sloth bear, Melursus ursinus, Ursus ursinus -mongoose -meerkat, mierkat -tiger beetle -ladybug, ladybeetle, lady beetle, ladybird, ladybird beetle -ground beetle, carabid beetle -long-horned beetle, longicorn, longicorn beetle -leaf beetle, chrysomelid -dung beetle -rhinoceros beetle -weevil -fly -bee -ant, emmet, pismire -grasshopper, hopper -cricket -walking stick, walkingstick, stick insect -cockroach, roach -mantis, mantid -cicada, cicala -leafhopper -lacewing, lacewing fly -dragonfly, darning needle, devil's darning needle, sewing needle, snake feeder, snake doctor, mosquito hawk, skeeter hawk -damselfly -admiral -ringlet, ringlet butterfly -monarch, monarch butterfly, milkweed butterfly, Danaus plexippus -cabbage butterfly -sulphur butterfly, sulfur butterfly -lycaenid, lycaenid butterfly -starfish, sea star -sea urchin -sea cucumber, holothurian -wood rabbit, cottontail, cottontail rabbit -hare -Angora, Angora rabbit -hamster -porcupine, hedgehog -fox squirrel, eastern fox squirrel, Sciurus niger -marmot -beaver -guinea pig, Cavia cobaya -sorrel -zebra -hog, pig, grunter, squealer, Sus scrofa -wild boar, boar, Sus scrofa -warthog -hippopotamus, hippo, river horse, Hippopotamus amphibius -ox -water buffalo, water ox, Asiatic buffalo, Bubalus bubalis -bison -ram, tup -bighorn, bighorn sheep, cimarron, Rocky Mountain bighorn, Rocky Mountain sheep, Ovis canadensis -ibex, Capra ibex -hartebeest -impala, Aepyceros melampus -gazelle -Arabian camel, dromedary, Camelus dromedarius -llama -weasel -mink -polecat, fitch, foulmart, foumart, Mustela putorius -black-footed ferret, ferret, Mustela nigripes -otter -skunk, polecat, wood pussy -badger -armadillo -three-toed sloth, ai, Bradypus tridactylus -orangutan, orang, orangutang, Pongo pygmaeus -gorilla, Gorilla gorilla -chimpanzee, chimp, Pan troglodytes -gibbon, Hylobates lar -siamang, Hylobates syndactylus, Symphalangus syndactylus -guenon, guenon monkey -patas, hussar monkey, Erythrocebus patas -baboon -macaque -langur -colobus, colobus monkey -proboscis monkey, Nasalis larvatus -marmoset -capuchin, ringtail, Cebus capucinus -howler monkey, howler -titi, titi monkey -spider monkey, Ateles geoffroyi -squirrel monkey, Saimiri sciureus -Madagascar cat, ring-tailed lemur, Lemur catta -indri, indris, Indri indri, Indri brevicaudatus -Indian elephant, Elephas maximus -African elephant, Loxodonta africana -lesser panda, red panda, panda, bear cat, cat bear, Ailurus fulgens -giant panda, panda, panda bear, coon bear, Ailuropoda melanoleuca -barracouta, snoek -eel -coho, cohoe, coho salmon, blue jack, silver salmon, Oncorhynchus kisutch -rock beauty, Holocanthus tricolor -anemone fish -sturgeon -gar, garfish, garpike, billfish, Lepisosteus osseus -lionfish -puffer, pufferfish, blowfish, globefish -abacus -abaya -academic gown, academic robe, judge's robe -accordion, piano accordion, squeeze box -acoustic guitar -aircraft carrier, carrier, flattop, attack aircraft carrier -airliner -airship, dirigible -altar -ambulance -amphibian, amphibious vehicle -analog clock -apiary, bee house -apron -ashcan, trash can, garbage can, wastebin, ash bin, ash-bin, ashbin, dustbin, trash barrel, trash bin -assault rifle, assault gun -backpack, back pack, knapsack, packsack, rucksack, haversack -bakery, bakeshop, bakehouse -balance beam, beam -balloon -ballpoint, ballpoint pen, ballpen, Biro -Band Aid -banjo -bannister, banister, balustrade, balusters, handrail -barbell -barber chair -barbershop -barn -barometer -barrel, cask -barrow, garden cart, lawn cart, wheelbarrow -baseball -basketball -bassinet -bassoon -bathing cap, swimming cap -bath towel -bathtub, bathing tub, bath, tub -beach wagon, station wagon, wagon, estate car, beach waggon, station waggon, waggon -beacon, lighthouse, beacon light, pharos -beaker -bearskin, busby, shako -beer bottle -beer glass -bell cote, bell cot -bib -bicycle-built-for-two, tandem bicycle, tandem -bikini, two-piece -binder, ring-binder -binoculars, field glasses, opera glasses -birdhouse -boathouse -bobsled, bobsleigh, bob -bolo tie, bolo, bola tie, bola -bonnet, poke bonnet -bookcase -bookshop, bookstore, bookstall -bottlecap -bow -bow tie, bow-tie, bowtie -brass, memorial tablet, plaque -brassiere, bra, bandeau -breakwater, groin, groyne, mole, bulwark, seawall, jetty -breastplate, aegis, egis -broom -bucket, pail -buckle -bulletproof vest -bullet train, bullet -butcher shop, meat market -cab, hack, taxi, taxicab -caldron, cauldron -candle, taper, wax light -cannon -canoe -can opener, tin opener -cardigan -car mirror -carousel, carrousel, merry-go-round, roundabout, whirligig -carpenter's kit, tool kit -carton -car wheel -cash machine, cash dispenser, automated teller machine, automatic teller machine, automated teller, automatic teller, ATM -cassette -cassette player -castle -catamaran -CD player -cello, violoncello -cellular telephone, cellular phone, cellphone, cell, mobile phone -chain -chainlink fence -chain mail, ring mail, mail, chain armor, chain armour, ring armor, ring armour -chain saw, chainsaw -chest -chiffonier, commode -chime, bell, gong -china cabinet, china closet -Christmas stocking -church, church building -cinema, movie theater, movie theatre, movie house, picture palace -cleaver, meat cleaver, chopper -cliff dwelling -cloak -clog, geta, patten, sabot -cocktail shaker -coffee mug -coffeepot -coil, spiral, volute, whorl, helix -combination lock -computer keyboard, keypad -confectionery, confectionary, candy store -container ship, containership, container vessel -convertible -corkscrew, bottle screw -cornet, horn, trumpet, trump -cowboy boot -cowboy hat, ten-gallon hat -cradle -crane -crash helmet -crate -crib, cot -Crock Pot -croquet ball -crutch -cuirass -dam, dike, dyke -desk -desktop computer -dial telephone, dial phone -diaper, nappy, napkin -digital clock -digital watch -dining table, board -dishrag, dishcloth -dishwasher, dish washer, dishwashing machine -disk brake, disc brake -dock, dockage, docking facility -dogsled, dog sled, dog sleigh -dome -doormat, welcome mat -drilling platform, offshore rig -drum, membranophone, tympan -drumstick -dumbbell -Dutch oven -electric fan, blower -electric guitar -electric locomotive -entertainment center -envelope -espresso maker -face powder -feather boa, boa -file, file cabinet, filing cabinet -fireboat -fire engine, fire truck -fire screen, fireguard -flagpole, flagstaff -flute, transverse flute -folding chair -football helmet -forklift -fountain -fountain pen -four-poster -freight car -French horn, horn -frying pan, frypan, skillet -fur coat -garbage truck, dustcart -gasmask, respirator, gas helmet -gas pump, gasoline pump, petrol pump, island dispenser -goblet -go-kart -golf ball -golfcart, golf cart -gondola -gong, tam-tam -gown -grand piano, grand -greenhouse, nursery, glasshouse -grille, radiator grille -grocery store, grocery, food market, market -guillotine -hair slide -hair spray -half track -hammer -hamper -hand blower, blow dryer, blow drier, hair dryer, hair drier -hand-held computer, hand-held microcomputer -handkerchief, hankie, hanky, hankey -hard disc, hard disk, fixed disk -harmonica, mouth organ, harp, mouth harp -harp -harvester, reaper -hatchet -holster -home theater, home theatre -honeycomb -hook, claw -hoopskirt, crinoline -horizontal bar, high bar -horse cart, horse-cart -hourglass -iPod -iron, smoothing iron -jack-o'-lantern -jean, blue jean, denim -jeep, landrover -jersey, T-shirt, tee shirt -jigsaw puzzle -jinrikisha, ricksha, rickshaw -joystick -kimono -knee pad -knot -lab coat, laboratory coat -ladle -lampshade, lamp shade -laptop, laptop computer -lawn mower, mower -lens cap, lens cover -letter opener, paper knife, paperknife -library -lifeboat -lighter, light, igniter, ignitor -limousine, limo -liner, ocean liner -lipstick, lip rouge -Loafer -lotion -loudspeaker, speaker, speaker unit, loudspeaker system, speaker system -loupe, jeweler's loupe -lumbermill, sawmill -magnetic compass -mailbag, postbag -mailbox, letter box -maillot -maillot, tank suit -manhole cover -maraca -marimba, xylophone -mask -matchstick -maypole -maze, labyrinth -measuring cup -medicine chest, medicine cabinet -megalith, megalithic structure -microphone, mike -microwave, microwave oven -military uniform -milk can -minibus -miniskirt, mini -minivan -missile -mitten -mixing bowl -mobile home, manufactured home -Model T -modem -monastery -monitor -moped -mortar -mortarboard -mosque -mosquito net -motor scooter, scooter -mountain bike, all-terrain bike, off-roader -mountain tent -mouse, computer mouse -mousetrap -moving van -muzzle -nail -neck brace -necklace -nipple -notebook, notebook computer -obelisk -oboe, hautboy, hautbois -ocarina, sweet potato -odometer, hodometer, mileometer, milometer -oil filter -organ, pipe organ -oscilloscope, scope, cathode-ray oscilloscope, CRO -overskirt -oxcart -oxygen mask -packet -paddle, boat paddle -paddlewheel, paddle wheel -padlock -paintbrush -pajama, pyjama, pj's, jammies -palace -panpipe, pandean pipe, syrinx -paper towel -parachute, chute -parallel bars, bars -park bench -parking meter -passenger car, coach, carriage -patio, terrace -pay-phone, pay-station -pedestal, plinth, footstall -pencil box, pencil case -pencil sharpener -perfume, essence -Petri dish -photocopier -pick, plectrum, plectron -pickelhaube -picket fence, paling -pickup, pickup truck -pier -piggy bank, penny bank -pill bottle -pillow -ping-pong ball -pinwheel -pirate, pirate ship -pitcher, ewer -plane, carpenter's plane, woodworking plane -planetarium -plastic bag -plate rack -plow, plough -plunger, plumber's helper -Polaroid camera, Polaroid Land camera -pole -police van, police wagon, paddy wagon, patrol wagon, wagon, black Maria -poncho -pool table, billiard table, snooker table -pop bottle, soda bottle -pot, flowerpot -potter's wheel -power drill -prayer rug, prayer mat -printer -prison, prison house -projectile, missile -projector -puck, hockey puck -punching bag, punch bag, punching ball, punchball -purse -quill, quill pen -quilt, comforter, comfort, puff -racer, race car, racing car -racket, racquet -radiator -radio, wireless -radio telescope, radio reflector -rain barrel -recreational vehicle, RV, R.V. -reel -reflex camera -refrigerator, icebox -remote control, remote -restaurant, eating house, eating place, eatery -revolver, six-gun, six-shooter -rifle -rocking chair, rocker -rotisserie -rubber eraser, rubber, pencil eraser -rugby ball -rule, ruler -running shoe -safe -safety pin -saltshaker, salt shaker -sandal -sarong -sax, saxophone -scabbard -scale, weighing machine -school bus -schooner -scoreboard -screen, CRT screen -screw -screwdriver -seat belt, seatbelt -sewing machine -shield, buckler -shoe shop, shoe-shop, shoe store -shoji -shopping basket -shopping cart -shovel -shower cap -shower curtain -ski -ski mask -sleeping bag -slide rule, slipstick -sliding door -slot, one-armed bandit -snorkel -snowmobile -snowplow, snowplough -soap dispenser -soccer ball -sock -solar dish, solar collector, solar furnace -sombrero -soup bowl -space bar -space heater -space shuttle -spatula -speedboat -spider web, spider's web -spindle -sports car, sport car -spotlight, spot -stage -steam locomotive -steel arch bridge -steel drum -stethoscope -stole -stone wall -stopwatch, stop watch -stove -strainer -streetcar, tram, tramcar, trolley, trolley car -stretcher -studio couch, day bed -stupa, tope -submarine, pigboat, sub, U-boat -suit, suit of clothes -sundial -sunglass -sunglasses, dark glasses, shades -sunscreen, sunblock, sun blocker -suspension bridge -swab, swob, mop -sweatshirt -swimming trunks, bathing trunks -swing -switch, electric switch, electrical switch -syringe -table lamp -tank, army tank, armored combat vehicle, armoured combat vehicle -tape player -teapot -teddy, teddy bear -television, television system -tennis ball -thatch, thatched roof -theater curtain, theatre curtain -thimble -thresher, thrasher, threshing machine -throne -tile roof -toaster -tobacco shop, tobacconist shop, tobacconist -toilet seat -torch -totem pole -tow truck, tow car, wrecker -toyshop -tractor -trailer truck, tractor trailer, trucking rig, rig, articulated lorry, semi -tray -trench coat -tricycle, trike, velocipede -trimaran -tripod -triumphal arch -trolleybus, trolley coach, trackless trolley -trombone -tub, vat -turnstile -typewriter keyboard -umbrella -unicycle, monocycle -upright, upright piano -vacuum, vacuum cleaner -vase -vault -velvet -vending machine -vestment -viaduct -violin, fiddle -volleyball -waffle iron -wall clock -wallet, billfold, notecase, pocketbook -wardrobe, closet, press -warplane, military plane -washbasin, handbasin, washbowl, lavabo, wash-hand basin -washer, automatic washer, washing machine -water bottle -water jug -water tower -whiskey jug -whistle -wig -window screen -window shade -Windsor tie -wine bottle -wing -wok -wooden spoon -wool, woolen, woollen -worm fence, snake fence, snake-rail fence, Virginia fence -wreck -yawl -yurt -web site, website, internet site, site -comic book -crossword puzzle, crossword -street sign -traffic light, traffic signal, stoplight -book jacket, dust cover, dust jacket, dust wrapper -menu -plate -guacamole -consomme -hot pot, hotpot -trifle -ice cream, icecream -ice lolly, lolly, lollipop, popsicle -French loaf -bagel, beigel -pretzel -cheeseburger -hotdog, hot dog, red hot -mashed potato -head cabbage -broccoli -cauliflower -zucchini, courgette -spaghetti squash -acorn squash -butternut squash -cucumber, cuke -artichoke, globe artichoke -bell pepper -cardoon -mushroom -Granny Smith -strawberry -orange -lemon -fig -pineapple, ananas -banana -jackfruit, jak, jack -custard apple -pomegranate -hay -carbonara -chocolate sauce, chocolate syrup -dough -meat loaf, meatloaf -pizza, pizza pie -potpie -burrito -red wine -espresso -cup -eggnog -alp -bubble -cliff, drop, drop-off -coral reef -geyser -lakeside, lakeshore -promontory, headland, head, foreland -sandbar, sand bar -seashore, coast, seacoast, sea-coast -valley, vale -volcano -ballplayer, baseball player -groom, bridegroom -scuba diver -rapeseed -daisy -yellow lady's slipper, yellow lady-slipper, Cypripedium calceolus, Cypripedium parviflorum -corn -acorn -hip, rose hip, rosehip -buckeye, horse chestnut, conker -coral fungus -agaric -gyromitra -stinkhorn, carrion fungus -earthstar -hen-of-the-woods, hen of the woods, Polyporus frondosus, Grifola frondosa -bolete -ear, spike, capitulum -toilet tissue, toilet paper, bathroom tissue diff --git a/hub_module/modules/image/classification/efficientnetb4_imagenet/layers.py b/hub_module/modules/image/classification/efficientnetb4_imagenet/layers.py deleted file mode 100644 index 00f792b5..00000000 --- a/hub_module/modules/image/classification/efficientnetb4_imagenet/layers.py +++ /dev/null @@ -1,248 +0,0 @@ -# -*- coding:utf-8 -*- -# copyright (c) 2020 PaddlePaddle Authors. All Rights Reserve. -# -# Licensed under the Apache License, Version 2.0 (the "License"); -# you may not use this file except in compliance with the License. -# You may obtain a copy of the License at -# -# http://www.apache.org/licenses/LICENSE-2.0 -# -# Unless required by applicable law or agreed to in writing, software -# distributed under the License is distributed on an "AS IS" BASIS, -# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. -# See the License for the specific language governing permissions and -# limitations under the License. - -from __future__ import absolute_import -from __future__ import division -from __future__ import print_function - -import math -import warnings - -import paddle.fluid as fluid - - -def initial_type(name, - input, - op_type, - fan_out, - init="google", - use_bias=False, - filter_size=0, - stddev=0.02): - if init == "kaiming": - if op_type == 'conv': - fan_in = input.shape[1] * filter_size * filter_size - elif op_type == 'deconv': - fan_in = fan_out * filter_size * filter_size - else: - if len(input.shape) > 2: - fan_in = input.shape[1] * input.shape[2] * input.shape[3] - else: - fan_in = input.shape[1] - bound = 1 / math.sqrt(fan_in) - param_attr = fluid.ParamAttr( - name=name + "_weights", - initializer=fluid.initializer.Uniform(low=-bound, high=bound)) - if use_bias == True: - bias_attr = fluid.ParamAttr( - name=name + '_offset', - initializer=fluid.initializer.Uniform(low=-bound, high=bound)) - else: - bias_attr = False - elif init == 'google': - n = filter_size * filter_size * fan_out - param_attr = fluid.ParamAttr( - name=name + "_weights", - initializer=fluid.initializer.NormalInitializer( - loc=0.0, scale=math.sqrt(2.0 / n))) - if use_bias == True: - bias_attr = fluid.ParamAttr( - name=name + "_offset", - initializer=fluid.initializer.Constant(0.0)) - else: - bias_attr = False - - else: - param_attr = fluid.ParamAttr( - name=name + "_weights", - initializer=fluid.initializer.NormalInitializer( - loc=0.0, scale=stddev)) - if use_bias == True: - bias_attr = fluid.ParamAttr( - name=name + "_offset", - initializer=fluid.initializer.Constant(0.0)) - else: - bias_attr = False - return param_attr, bias_attr - - -def cal_padding(img_size, stride, filter_size, dilation=1): - """Calculate padding size.""" - if img_size % stride == 0: - out_size = max(filter_size - stride, 0) - else: - out_size = max(filter_size - (img_size % stride), 0) - return out_size // 2, out_size - out_size // 2 - - -def init_batch_norm_layer(name="batch_norm"): - param_attr = fluid.ParamAttr( - name=name + '_scale', initializer=fluid.initializer.Constant(1.0)) - bias_attr = fluid.ParamAttr( - name=name + '_offset', - initializer=fluid.initializer.Constant(value=0.0)) - return param_attr, bias_attr - - -def init_fc_layer(fout, name='fc'): - n = fout # fan-out - init_range = 1.0 / math.sqrt(n) - - param_attr = fluid.ParamAttr( - name=name + '_weights', - initializer=fluid.initializer.UniformInitializer( - low=-init_range, high=init_range)) - bias_attr = fluid.ParamAttr( - name=name + '_offset', - initializer=fluid.initializer.Constant(value=0.0)) - return param_attr, bias_attr - - -def norm_layer(input, norm_type='batch_norm', name=None): - if norm_type == 'batch_norm': - param_attr = fluid.ParamAttr( - name=name + '_weights', initializer=fluid.initializer.Constant(1.0)) - bias_attr = fluid.ParamAttr( - name=name + '_offset', - initializer=fluid.initializer.Constant(value=0.0)) - return fluid.layers.batch_norm( - input, - param_attr=param_attr, - bias_attr=bias_attr, - moving_mean_name=name + '_mean', - moving_variance_name=name + '_variance') - - elif norm_type == 'instance_norm': - helper = fluid.layer_helper.LayerHelper("instance_norm", **locals()) - dtype = helper.input_dtype() - epsilon = 1e-5 - mean = fluid.layers.reduce_mean(input, dim=[2, 3], keep_dim=True) - var = fluid.layers.reduce_mean( - fluid.layers.square(input - mean), dim=[2, 3], keep_dim=True) - if name is not None: - scale_name = name + "_scale" - offset_name = name + "_offset" - scale_param = fluid.ParamAttr( - name=scale_name, - initializer=fluid.initializer.Constant(1.0), - trainable=True) - offset_param = fluid.ParamAttr( - name=offset_name, - initializer=fluid.initializer.Constant(0.0), - trainable=True) - scale = helper.create_parameter( - attr=scale_param, shape=input.shape[1:2], dtype=dtype) - offset = helper.create_parameter( - attr=offset_param, shape=input.shape[1:2], dtype=dtype) - - tmp = fluid.layers.elementwise_mul(x=(input - mean), y=scale, axis=1) - tmp = tmp / fluid.layers.sqrt(var + epsilon) - tmp = fluid.layers.elementwise_add(tmp, offset, axis=1) - return tmp - else: - raise NotImplementedError("norm tyoe: [%s] is not support" % norm_type) - - -def conv2d(input, - num_filters=64, - filter_size=7, - stride=1, - stddev=0.02, - padding=0, - groups=None, - name="conv2d", - norm=None, - act=None, - relufactor=0.0, - use_bias=False, - padding_type=None, - initial="normal", - use_cudnn=True): - - if padding != 0 and padding_type != None: - warnings.warn( - 'padding value and padding type are set in the same time, and the final padding width and padding height are computed by padding_type' - ) - - param_attr, bias_attr = initial_type( - name=name, - input=input, - op_type='conv', - fan_out=num_filters, - init=initial, - use_bias=use_bias, - filter_size=filter_size, - stddev=stddev) - - def get_padding(filter_size, stride=1, dilation=1): - padding = ((stride - 1) + dilation * (filter_size - 1)) // 2 - return padding - - need_crop = False - if padding_type == "SAME": - top_padding, bottom_padding = cal_padding(input.shape[2], stride, - filter_size) - left_padding, right_padding = cal_padding(input.shape[2], stride, - filter_size) - height_padding = bottom_padding - width_padding = right_padding - if top_padding != bottom_padding or left_padding != right_padding: - height_padding = top_padding + stride - width_padding = left_padding + stride - need_crop = True - padding = [height_padding, width_padding] - elif padding_type == "VALID": - height_padding = 0 - width_padding = 0 - padding = [height_padding, width_padding] - elif padding_type == "DYNAMIC": - padding = get_padding(filter_size, stride) - else: - padding = padding - - conv = fluid.layers.conv2d( - input, - num_filters, - filter_size, - groups=groups, - name=name, - stride=stride, - padding=padding, - use_cudnn=use_cudnn, - param_attr=param_attr, - bias_attr=bias_attr) - - if need_crop: - conv = conv[:, :, 1:, 1:] - - if norm is not None: - conv = norm_layer(input=conv, norm_type=norm, name=name + "_norm") - if act == 'relu': - conv = fluid.layers.relu(conv, name=name + '_relu') - elif act == 'leaky_relu': - conv = fluid.layers.leaky_relu( - conv, alpha=relufactor, name=name + '_leaky_relu') - elif act == 'tanh': - conv = fluid.layers.tanh(conv, name=name + '_tanh') - elif act == 'sigmoid': - conv = fluid.layers.sigmoid(conv, name=name + '_sigmoid') - elif act == 'swish': - conv = fluid.layers.swish(conv, name=name + '_swish') - elif act == None: - conv = conv - else: - raise NotImplementedError("activation: [%s] is not support" % act) - - return conv diff --git a/hub_module/modules/image/classification/efficientnetb4_imagenet/module.py b/hub_module/modules/image/classification/efficientnetb4_imagenet/module.py index 426f7176..1aebdf78 100644 --- a/hub_module/modules/image/classification/efficientnetb4_imagenet/module.py +++ b/hub_module/modules/image/classification/efficientnetb4_imagenet/module.py @@ -1,4 +1,3 @@ -# -*- coding:utf-8 -*- # copyright (c) 2020 PaddlePaddle Authors. All Rights Reserve. # # Licensed under the Apache License, Version 2.0 (the "License"); @@ -13,318 +12,747 @@ # See the License for the specific language governing permissions and # limitations under the License. -from __future__ import absolute_import -from __future__ import division - -import ast -import argparse import os - -import numpy as np -import paddle.fluid as fluid -import paddlehub as hub -from paddle.fluid.core import PaddleTensor, AnalysisConfig, create_paddle_predictor -from paddlehub.module.module import moduleinfo, runnable, serving -from paddlehub.common.paddle_helper import add_vars_prefix - -from efficientnetb4_imagenet.processor import postprocess, base64_to_cv2 -from efficientnetb4_imagenet.data_feed import reader -from efficientnetb4_imagenet.efficientnet import EfficientNetB4 - - -@moduleinfo( - name="efficientnetb4_imagenet", - type="CV/image_classification", - author="paddlepaddle", - author_email="paddle-dev@baidu.com", - summary= - "EfficientNetB4 is a image classfication model, this module is trained with imagenet datasets.", - version="1.1.0") -class EfficientNetB4ImageNet(hub.Module): - def _initialize(self): - self.default_pretrained_model_path = os.path.join( - self.directory, "efficientnetb4_imagenet_infer_model") - label_file = os.path.join(self.directory, "label_list.txt") - with open(label_file, 'r', encoding='utf-8') as file: - self.label_list = file.read().split("\n")[:-1] - self.classification = self.classify - self._set_config() - - def get_expected_image_width(self): - return 224 - - def get_expected_image_height(self): - return 224 - - def get_pretrained_images_mean(self): - im_mean = np.array([0.485, 0.456, 0.406]).reshape(1, 3) - return im_mean - - def get_pretrained_images_std(self): - im_std = np.array([0.229, 0.224, 0.225]).reshape(1, 3) - return im_std - - def _set_config(self): - """ - predictor config setting - """ - cpu_config = AnalysisConfig(self.default_pretrained_model_path) - cpu_config.disable_glog_info() - cpu_config.disable_gpu() - self.cpu_predictor = create_paddle_predictor(cpu_config) - - try: - _places = os.environ["CUDA_VISIBLE_DEVICES"] - int(_places[0]) - use_gpu = True - except: - use_gpu = False - if use_gpu: - gpu_config = AnalysisConfig(self.default_pretrained_model_path) - gpu_config.disable_glog_info() - gpu_config.enable_use_gpu( - memory_pool_init_size_mb=1000, device_id=0) - self.gpu_predictor = create_paddle_predictor(gpu_config) - - def context(self, - trainable=True, - pretrained=True, - override_params=None, - phase='train'): - """context for transfer learning. - - Args: - trainable (bool): Set parameters in program to be trainable. - pretrained (bool) : Whether to load pretrained model. - - Returns: - inputs (dict): key is 'image', corresponding vaule is image tensor. - outputs (dict): key is : - 'classification', corresponding value is the result of classification. - 'feature_map', corresponding value is the result of the layer before the fully connected layer. - context_prog (fluid.Program): program for transfer learning. - """ - if phase in ["dev", "test", "predict", "eval"]: - is_test = True - elif phase in ["train"]: - is_test = False - else: - raise ValueError( - "Phase %s is error, which must be one of train, dev, test, eval and predict." - % phase) - - context_prog = fluid.Program() - startup_prog = fluid.Program() - with fluid.program_guard(context_prog, startup_prog): - with fluid.unique_name.guard(): - image = fluid.layers.data( - name="image", shape=[3, 224, 224], dtype="float32") - efficientnet_b4 = EfficientNetB4( - override_params=override_params) - output, feature_map = efficientnet_b4.net( - input=image, - class_dim=len(self.label_list), - is_test=is_test) - - name_prefix = '@HUB_{}@'.format(self.name) - inputs = {'image': name_prefix + image.name} - outputs = { - 'classification': name_prefix + output.name, - 'feature_map': name_prefix + feature_map.name - } - add_vars_prefix(context_prog, name_prefix) - add_vars_prefix(startup_prog, name_prefix) - global_vars = context_prog.global_block().vars - inputs = { - key: global_vars[value] - for key, value in inputs.items() - } - outputs = { - key: global_vars[value] - for key, value in outputs.items() - } - - place = fluid.CPUPlace() - exe = fluid.Executor(place) - # pretrained - if pretrained: - - def _if_exist(var): - b = os.path.exists( - os.path.join(self.default_pretrained_model_path, - var.name)) - return b - - fluid.io.load_vars( - exe, - self.default_pretrained_model_path, - context_prog, - predicate=_if_exist) - else: - exe.run(startup_prog) - # trainable - for param in context_prog.global_block().iter_parameters(): - param.trainable = trainable - return inputs, outputs, context_prog - - def classify(self, - images=None, - paths=None, - batch_size=1, - use_gpu=False, - top_k=1): +import math +import collections +import re +import copy + +import paddle +from paddle import ParamAttr +import paddle.nn as nn +import paddle.nn.functional as F +from paddle.nn import Conv2d, BatchNorm, Linear, Dropout +from paddle.nn import AdaptiveAvgPool2d, MaxPool2d, AvgPool2d +from paddlehub.module.module import moduleinfo +from paddlehub.module.cv_module import ImageClassifierModule + +GlobalParams = collections.namedtuple('GlobalParams', [ + 'batch_norm_momentum', + 'batch_norm_epsilon', + 'dropout_rate', + 'num_classes', + 'width_coefficient', + 'depth_coefficient', + 'depth_divisor', + 'min_depth', + 'drop_connect_rate', +]) + +BlockArgs = collections.namedtuple( + 'BlockArgs', + ['kernel_size', 'num_repeat', 'input_filters', 'output_filters', 'expand_ratio', 'id_skip', 'stride', 'se_ratio']) + +GlobalParams.__new__.__defaults__ = (None, ) * len(GlobalParams._fields) +BlockArgs.__new__.__defaults__ = (None, ) * len(BlockArgs._fields) + + +def efficientnet_params(model_name: str): + """ Map EfficientNet model name to parameter coefficients. """ + params_dict = { + # Coefficients: width,depth,resolution,dropout + 'efficientnet-b4': (1.4, 1.8, 380, 0.4) + } + return params_dict[model_name] + + +def efficientnet(width_coefficient: float = None, + depth_coefficient: float = None, + dropout_rate: float = 0.2, + drop_connect_rate: float = 0.2): + """ Get block arguments according to parameter and coefficients. """ + blocks_args = [ + 'r1_k3_s11_e1_i32_o16_se0.25', + 'r2_k3_s22_e6_i16_o24_se0.25', + 'r2_k5_s22_e6_i24_o40_se0.25', + 'r3_k3_s22_e6_i40_o80_se0.25', + 'r3_k5_s11_e6_i80_o112_se0.25', + 'r4_k5_s22_e6_i112_o192_se0.25', + 'r1_k3_s11_e6_i192_o320_se0.25', + ] + blocks_args = BlockDecoder.decode(blocks_args) + + global_params = GlobalParams(batch_norm_momentum=0.99, + batch_norm_epsilon=1e-3, + dropout_rate=dropout_rate, + drop_connect_rate=drop_connect_rate, + num_classes=1000, + width_coefficient=width_coefficient, + depth_coefficient=depth_coefficient, + depth_divisor=8, + min_depth=None) + + return blocks_args, global_params + + +def get_model_params(model_name: str, override_params: dict): + """ Get the block args and global params for a given model """ + if model_name.startswith('efficientnet'): + w, d, _, p = efficientnet_params(model_name) + blocks_args, global_params = efficientnet(width_coefficient=w, depth_coefficient=d, dropout_rate=p) + else: + raise NotImplementedError('model name is not pre-defined: %s' % model_name) + if override_params: + global_params = global_params._replace(**override_params) + return blocks_args, global_params + + +def round_filters(filters: int, global_params: dict): + """ Calculate and round number of filters based on depth multiplier. """ + multiplier = global_params.width_coefficient + if not multiplier: + return filters + divisor = global_params.depth_divisor + min_depth = global_params.min_depth + filters *= multiplier + min_depth = min_depth or divisor + new_filters = max(min_depth, int(filters + divisor / 2) // divisor * divisor) + if new_filters < 0.9 * filters: # prevent rounding by more than 10% + new_filters += divisor + return int(new_filters) + + +def round_repeats(repeats: int, global_params: dict): + """ Round number of filters based on depth multiplier. """ + multiplier = global_params.depth_coefficient + if not multiplier: + return repeats + return int(math.ceil(multiplier * repeats)) + + +class BlockDecoder(object): + """ + Block Decoder, straight from the official TensorFlow repository. + """ + @staticmethod + def _decode_block_string(block_string: str): + """ Gets a block through a string notation of arguments. """ + assert isinstance(block_string, str) + + ops = block_string.split('_') + options = {} + for op in ops: + splits = re.split(r'(\d.*)', op) + if len(splits) >= 2: + key, value = splits[:2] + options[key] = value + + # Check stride + cond_1 = ('s' in options and len(options['s']) == 1) + cond_2 = ((len(options['s']) == 2) and (options['s'][0] == options['s'][1])) + assert (cond_1 or cond_2) + + return BlockArgs(kernel_size=int(options['k']), + num_repeat=int(options['r']), + input_filters=int(options['i']), + output_filters=int(options['o']), + expand_ratio=int(options['e']), + id_skip=('noskip' not in block_string), + se_ratio=float(options['se']) if 'se' in options else None, + stride=[int(options['s'][0])]) + + @staticmethod + def _encode_block_string(block): + """Encodes a block to a string.""" + args = [ + 'r%d' % block.num_repeat, + 'k%d' % block.kernel_size, + 's%d%d' % (block.strides[0], block.strides[1]), + 'e%s' % block.expand_ratio, + 'i%d' % block.input_filters, + 'o%d' % block.output_filters + ] + if 0 < block.se_ratio <= 1: + args.append('se%s' % block.se_ratio) + if block.id_skip is False: + args.append('noskip') + return '_'.join(args) + + @staticmethod + def decode(string_list: list): """ - API for image classification. - - Args: - images (list[numpy.ndarray]): data of images, shape of each is [H, W, C], color space must be BGR. - paths (list[str]): The paths of images. - batch_size (int): batch size. - use_gpu (bool): Whether to use gpu. - top_k (int): Return top k results. + Decode a list of string notations to specify blocks in the network. - Returns: - res (list[dict]): The classfication results. + string_list: list of strings, each string is a notation of block + return + list of BlockArgs namedtuples of block args """ - if use_gpu: - try: - _places = os.environ["CUDA_VISIBLE_DEVICES"] - int(_places[0]) - except: - raise RuntimeError( - "Environment Variable CUDA_VISIBLE_DEVICES is not set correctly. If you wanna use gpu, please set CUDA_VISIBLE_DEVICES as cuda_device_id." - ) - - all_data = list() - for yield_data in reader(images, paths): - all_data.append(yield_data) - - total_num = len(all_data) - loop_num = int(np.ceil(total_num / batch_size)) - - res = list() - for iter_id in range(loop_num): - batch_data = list() - handle_id = iter_id * batch_size - for image_id in range(batch_size): - try: - batch_data.append(all_data[handle_id + image_id]) - except: - pass - # feed batch image - batch_image = np.array([data['image'] for data in batch_data]) - batch_image = PaddleTensor(batch_image.copy()) - predictor_output = self.gpu_predictor.run([ - batch_image - ]) if use_gpu else self.cpu_predictor.run([batch_image]) - out = postprocess( - data_out=predictor_output[0].as_ndarray(), - label_list=self.label_list, - top_k=top_k) - res += out - return res - - def save_inference_model(self, - dirname, - model_filename=None, - params_filename=None, - combined=True): - if combined: - model_filename = "__model__" if not model_filename else model_filename - params_filename = "__params__" if not params_filename else params_filename - place = fluid.CPUPlace() - exe = fluid.Executor(place) - - program, feeded_var_names, target_vars = fluid.io.load_inference_model( - dirname=self.default_pretrained_model_path, executor=exe) - - fluid.io.save_inference_model( - dirname=dirname, - main_program=program, - executor=exe, - feeded_var_names=feeded_var_names, - target_vars=target_vars, - model_filename=model_filename, - params_filename=params_filename) - - @serving - def serving_method(self, images, **kwargs): + assert isinstance(string_list, list) + blocks_args = [] + for block_string in string_list: + blocks_args.append(BlockDecoder._decode_block_string(block_string)) + return blocks_args + + @staticmethod + def encode(blocks_args: list): """ - Run as a service. - """ - images_decode = [base64_to_cv2(image) for image in images] - results = self.classify(images=images_decode, **kwargs) - return results + Encodes a list of BlockArgs to a list of strings. - @runnable - def run_cmd(self, argvs): - """ - Run as a command. + :param blocks_args: a list of BlockArgs namedtuples of block args + :return: a list of strings, each string is a notation of block """ - self.parser = argparse.ArgumentParser( - description="Run the {} module.".format(self.name), - prog='hub run {}'.format(self.name), - usage='%(prog)s', - add_help=True) - self.arg_input_group = self.parser.add_argument_group( - title="Input options", description="Input data. Required") - self.arg_config_group = self.parser.add_argument_group( - title="Config options", - description= - "Run configuration for controlling module behavior, not required.") - self.add_module_config_arg() - self.add_module_input_arg() - args = self.parser.parse_args(argvs) - results = self.classify( - paths=[args.input_path], - batch_size=args.batch_size, - use_gpu=args.use_gpu) - return results - - def add_module_config_arg(self): - """ - Add the command config options. - """ - self.arg_config_group.add_argument( - '--use_gpu', - type=ast.literal_eval, - default=False, - help="whether use GPU or not.") - self.arg_config_group.add_argument( - '--batch_size', - type=ast.literal_eval, - default=1, - help="batch size.") - self.arg_config_group.add_argument( - '--top_k', - type=ast.literal_eval, - default=1, - help="Return top k results.") - - def add_module_input_arg(self): - """ - Add the command input options. - """ - self.arg_input_group.add_argument( - '--input_path', type=str, help="path to image.") - - -if __name__ == '__main__': - b4 = EfficientNetB4ImageNet() - b4.context() - import cv2 - test_image = [ - cv2.imread( - '/mnt/zhangxuefei/program-paddle/PaddleHub/hub_module/tests/image_dataset/classification/animals/dog.jpeg' - ) - ] - res = b4.classification(images=test_image) - print(res) - res = b4.classification(paths=[ - '/mnt/zhangxuefei/program-paddle/PaddleHub/hub_module/tests/image_dataset/classification/animals/dog.jpeg' - ]) - print(res) - res = b4.classification(images=test_image) - print(res) - res = b4.classify(images=test_image) - print(res) + block_strings = [] + for block in blocks_args: + block_strings.append(BlockDecoder._encode_block_string(block)) + return block_strings + + +def initial_type(name: str, use_bias: bool = False): + param_attr = ParamAttr(name=name + "_weights") + if use_bias: + bias_attr = ParamAttr(name=name + "_offset") + else: + bias_attr = False + return param_attr, bias_attr + + +def init_batch_norm_layer(name: str = "batch_norm"): + param_attr = ParamAttr(name=name + "_scale") + bias_attr = ParamAttr(name=name + "_offset") + return param_attr, bias_attr + + +def init_fc_layer(name: str = "fc"): + param_attr = ParamAttr(name=name + "_weights") + bias_attr = ParamAttr(name=name + "_offset") + return param_attr, bias_attr + + +def cal_padding(img_size: int, stride: int, filter_size: int, dilation: int = 1): + """Calculate padding size.""" + if img_size % stride == 0: + out_size = max(filter_size - stride, 0) + else: + out_size = max(filter_size - (img_size % stride), 0) + return out_size // 2, out_size - out_size // 2 + + +inp_shape = {"b4": [380, 190, 190, 95, 48, 24, 24, 12]} + + +def _drop_connect(inputs: paddle.Tensor, prob: float, is_test: bool): + """Drop input connection""" + if is_test: + return inputs + keep_prob = 1.0 - prob + inputs_shape = paddle.shape(inputs) + random_tensor = keep_prob + paddle.rand(shape=[inputs_shape[0], 1, 1, 1]) + binary_tensor = paddle.floor(random_tensor) + output = inputs / keep_prob * binary_tensor + return output + + +class Conv2ds(nn.Layer): + """Basic conv layer""" + def __init__(self, + input_channels: int, + output_channels: int, + filter_size: int, + stride: int = 1, + padding: int = 0, + groups: int = None, + name: str = "conv2d", + act: str = None, + use_bias: bool = False, + padding_type: str = None, + model_name: str = None, + cur_stage: str = None): + super(Conv2ds, self).__init__() + assert act in [None, "swish", "sigmoid"] + self.act = act + + param_attr, bias_attr = initial_type(name=name, use_bias=use_bias) + + def get_padding(filter_size, stride=1, dilation=1): + padding = ((stride - 1) + dilation * (filter_size - 1)) // 2 + return padding + + inps = 1 if model_name == None and cur_stage == None else inp_shape[model_name][cur_stage] + self.need_crop = False + if padding_type == "SAME": + top_padding, bottom_padding = cal_padding(inps, stride, filter_size) + left_padding, right_padding = cal_padding(inps, stride, filter_size) + height_padding = bottom_padding + width_padding = right_padding + if top_padding != bottom_padding or left_padding != right_padding: + height_padding = top_padding + stride + width_padding = left_padding + stride + self.need_crop = True + padding = [height_padding, width_padding] + elif padding_type == "VALID": + height_padding = 0 + width_padding = 0 + padding = [height_padding, width_padding] + elif padding_type == "DYNAMIC": + padding = get_padding(filter_size, stride) + else: + padding = padding_type + + groups = 1 if groups is None else groups + self._conv = Conv2d(input_channels, + output_channels, + filter_size, + groups=groups, + stride=stride, + padding=padding, + weight_attr=param_attr, + bias_attr=bias_attr) + + def forward(self, inputs: paddle.Tensor): + x = self._conv(inputs) + if self.act == "swish": + x = F.swish(x) + elif self.act == "sigmoid": + x = F.sigmoid(x) + + if self.need_crop: + x = x[:, :, 1:, 1:] + return x + + +class ConvBNLayer(nn.Layer): + """Basic conv bn layer.""" + def __init__(self, + input_channels: int, + filter_size: int, + output_channels: int, + stride: int = 1, + num_groups: int = 1, + padding_type: str = "SAME", + conv_act: str = None, + bn_act: str = "swish", + use_bn: bool = True, + use_bias: bool = False, + name: str = None, + conv_name: str = None, + bn_name: str = None, + model_name: str = None, + cur_stage: str = None): + super(ConvBNLayer, self).__init__() + + self._conv = Conv2ds(input_channels=input_channels, + output_channels=output_channels, + filter_size=filter_size, + stride=stride, + groups=num_groups, + act=conv_act, + padding_type=padding_type, + name=conv_name, + use_bias=use_bias, + model_name=model_name, + cur_stage=cur_stage) + self.use_bn = use_bn + if use_bn is True: + bn_name = name + bn_name + param_attr, bias_attr = init_batch_norm_layer(bn_name) + + self._bn = BatchNorm(num_channels=output_channels, + act=bn_act, + momentum=0.99, + epsilon=0.001, + moving_mean_name=bn_name + "_mean", + moving_variance_name=bn_name + "_variance", + param_attr=param_attr, + bias_attr=bias_attr) + + def forward(self, inputs: paddle.Tensor): + if self.use_bn: + x = self._conv(inputs) + x = self._bn(x) + return x + else: + return self._conv(inputs) + + +class ExpandConvNorm(nn.Layer): + """Expand conv norm layer.""" + def __init__(self, + input_channels: int, + block_args: dict, + padding_type: str, + name: str = None, + model_name: str = None, + cur_stage: str = None): + super(ExpandConvNorm, self).__init__() + + self.oup = block_args.input_filters * block_args.expand_ratio + self.expand_ratio = block_args.expand_ratio + + if self.expand_ratio != 1: + self._conv = ConvBNLayer(input_channels, + 1, + self.oup, + bn_act=None, + padding_type=padding_type, + name=name, + conv_name=name + "_expand_conv", + bn_name="_bn0", + model_name=model_name, + cur_stage=cur_stage) + + def forward(self, inputs: paddle.Tensor): + if self.expand_ratio != 1: + return self._conv(inputs) + else: + return inputs + + +class DepthwiseConvNorm(nn.Layer): + """Depthwise conv norm layer.""" + def __init__(self, + input_channels: int, + block_args: dict, + padding_type: str, + name: str = None, + model_name: str = None, + cur_stage: str = None): + super(DepthwiseConvNorm, self).__init__() + + self.k = block_args.kernel_size + self.s = block_args.stride + if isinstance(self.s, list) or isinstance(self.s, tuple): + self.s = self.s[0] + oup = block_args.input_filters * block_args.expand_ratio + + self._conv = ConvBNLayer(input_channels, + self.k, + oup, + self.s, + num_groups=input_channels, + bn_act=None, + padding_type=padding_type, + name=name, + conv_name=name + "_depthwise_conv", + bn_name="_bn1", + model_name=model_name, + cur_stage=cur_stage) + + def forward(self, inputs: paddle.Tensor): + return self._conv(inputs) + + +class ProjectConvNorm(nn.Layer): + """Projection conv bn layer.""" + def __init__(self, + input_channels: int, + block_args: dict, + padding_type: str, + name: str = None, + model_name: str = None, + cur_stage: str = None): + super(ProjectConvNorm, self).__init__() + + final_oup = block_args.output_filters + + self._conv = ConvBNLayer(input_channels, + 1, + final_oup, + bn_act=None, + padding_type=padding_type, + name=name, + conv_name=name + "_project_conv", + bn_name="_bn2", + model_name=model_name, + cur_stage=cur_stage) + + def forward(self, inputs: paddle.Tensor): + return self._conv(inputs) + + +class SEBlock(nn.Layer): + """Basic Squeeze-and-Excitation block for Efficientnet.""" + def __init__(self, + input_channels: int, + num_squeezed_channels: int, + oup: int, + padding_type: str, + name: str = None, + model_name: str = None, + cur_stage: str = None): + super(SEBlock, self).__init__() + + self._pool = AdaptiveAvgPool2d(1) + self._conv1 = Conv2ds(input_channels, + num_squeezed_channels, + 1, + use_bias=True, + padding_type=padding_type, + act="swish", + name=name + "_se_reduce") + + self._conv2 = Conv2ds(num_squeezed_channels, + oup, + 1, + act="sigmoid", + use_bias=True, + padding_type=padding_type, + name=name + "_se_expand") + + def forward(self, inputs: paddle.Tensor): + x = self._pool(inputs) + x = self._conv1(x) + x = self._conv2(x) + return paddle.multiply(inputs, x) + + +class MbConvBlock(nn.Layer): + """Mobile inverted bottleneck convolution for Efficientnet.""" + def __init__(self, + input_channels: int, + block_args: dict, + padding_type: str, + use_se: bool, + name: str = None, + drop_connect_rate: float = None, + is_test: bool = False, + model_name: str = None, + cur_stage: str = None): + super(MbConvBlock, self).__init__() + + oup = block_args.input_filters * block_args.expand_ratio + self.block_args = block_args + self.has_se = use_se and (block_args.se_ratio is not None) and (0 < block_args.se_ratio <= 1) + self.id_skip = block_args.id_skip + self.expand_ratio = block_args.expand_ratio + self.drop_connect_rate = drop_connect_rate + self.is_test = is_test + + if self.expand_ratio != 1: + self._ecn = ExpandConvNorm(input_channels, + block_args, + padding_type=padding_type, + name=name, + model_name=model_name, + cur_stage=cur_stage) + + self._dcn = DepthwiseConvNorm(input_channels * block_args.expand_ratio, + block_args, + padding_type=padding_type, + name=name, + model_name=model_name, + cur_stage=cur_stage) + + if self.has_se: + num_squeezed_channels = max(1, int(block_args.input_filters * block_args.se_ratio)) + self._se = SEBlock(input_channels * block_args.expand_ratio, + num_squeezed_channels, + oup, + padding_type=padding_type, + name=name, + model_name=model_name, + cur_stage=cur_stage) + + self._pcn = ProjectConvNorm(input_channels * block_args.expand_ratio, + block_args, + padding_type=padding_type, + name=name, + model_name=model_name, + cur_stage=cur_stage) + + def forward(self, inputs: paddle.Tensor): + x = inputs + if self.expand_ratio != 1: + x = self._ecn(x) + x = F.swish(x) + x = self._dcn(x) + x = F.swish(x) + if self.has_se: + x = self._se(x) + x = self._pcn(x) + if self.id_skip and \ + self.block_args.stride == 1 and \ + self.block_args.input_filters == self.block_args.output_filters: + if self.drop_connect_rate: + x = _drop_connect(x, self.drop_connect_rate, self.is_test) + x = paddle.elementwise_add(x, inputs) + return x + + +class ConvStemNorm(nn.Layer): + """Basic conv stem norm block for extracting features.""" + def __init__(self, + input_channels: int, + padding_type: str, + _global_params: dict, + name: str = None, + model_name: str = None, + cur_stage: str = None): + super(ConvStemNorm, self).__init__() + + output_channels = round_filters(32, _global_params) + self._conv = ConvBNLayer(input_channels, + filter_size=3, + output_channels=output_channels, + stride=2, + bn_act=None, + padding_type=padding_type, + name="", + conv_name="_conv_stem", + bn_name="_bn0", + model_name=model_name, + cur_stage=cur_stage) + + def forward(self, inputs: paddle.Tensor): + return self._conv(inputs) + + +class ExtractFeatures(nn.Layer): + """Extract features.""" + def __init__(self, + input_channels: int, + _block_args: dict, + _global_params: dict, + padding_type: str, + use_se: bool, + is_test: bool, + model_name: str = None): + super(ExtractFeatures, self).__init__() + + self._global_params = _global_params + + self._conv_stem = ConvStemNorm(input_channels, + padding_type=padding_type, + _global_params=_global_params, + model_name=model_name, + cur_stage=0) + + self.block_args_copy = copy.deepcopy(_block_args) + idx = 0 + block_size = 0 + for block_arg in self.block_args_copy: + block_arg = block_arg._replace(input_filters=round_filters(block_arg.input_filters, _global_params), + output_filters=round_filters(block_arg.output_filters, _global_params), + num_repeat=round_repeats(block_arg.num_repeat, _global_params)) + block_size += 1 + for _ in range(block_arg.num_repeat - 1): + block_size += 1 + + self.conv_seq = [] + cur_stage = 1 + for block_args in _block_args: + block_args = block_args._replace(input_filters=round_filters(block_args.input_filters, _global_params), + output_filters=round_filters(block_args.output_filters, _global_params), + num_repeat=round_repeats(block_args.num_repeat, _global_params)) + + drop_connect_rate = self._global_params.drop_connect_rate if not is_test else 0 + if drop_connect_rate: + drop_connect_rate *= float(idx) / block_size + + _mc_block = self.add_sublayer( + "_blocks." + str(idx) + ".", + MbConvBlock(block_args.input_filters, + block_args=block_args, + padding_type=padding_type, + use_se=use_se, + name="_blocks." + str(idx) + ".", + drop_connect_rate=drop_connect_rate, + model_name=model_name, + cur_stage=cur_stage)) + self.conv_seq.append(_mc_block) + idx += 1 + if block_args.num_repeat > 1: + block_args = block_args._replace(input_filters=block_args.output_filters, stride=1) + for _ in range(block_args.num_repeat - 1): + drop_connect_rate = self._global_params.drop_connect_rate if not is_test else 0 + if drop_connect_rate: + drop_connect_rate *= float(idx) / block_size + _mc_block = self.add_sublayer( + "block." + str(idx) + ".", + MbConvBlock(block_args.input_filters, + block_args, + padding_type=padding_type, + use_se=use_se, + name="_blocks." + str(idx) + ".", + drop_connect_rate=drop_connect_rate, + model_name=model_name, + cur_stage=cur_stage)) + self.conv_seq.append(_mc_block) + idx += 1 + cur_stage += 1 + + def forward(self, inputs: paddle.Tensor): + x = self._conv_stem(inputs) + x = F.swish(x) + for _mc_block in self.conv_seq: + x = _mc_block(x) + return x + + +@moduleinfo(name="efficientnetb4_imagenet", + type="cv/classification", + author="paddlepaddle", + author_email="", + summary="efficientnetb4_imagenet is a classification model, " + "this module is trained with Imagenet dataset.", + version="1.1.0", + meta=ImageClassifierModule) +class EfficientNet_B4(nn.Layer): + def __init__(self, + is_test: bool = False, + padding_type: str = "SAME", + override_params: dict = None, + use_se: bool = True, + class_dim: int = 1000, + load_checkpoint: str = None): + super(EfficientNet_B4, self).__init__() + + model_name = 'efficientnet-b4' + self.name = "b4" + self._block_args, self._global_params = get_model_params(model_name, override_params) + self.padding_type = padding_type + self.use_se = use_se + self.is_test = is_test + + self._ef = ExtractFeatures(3, + self._block_args, + self._global_params, + self.padding_type, + self.use_se, + self.is_test, + model_name=self.name) + + output_channels = round_filters(1280, self._global_params) + oup = 448 + + self._conv = ConvBNLayer(oup, + 1, + output_channels, + bn_act="swish", + padding_type=self.padding_type, + name="", + conv_name="_conv_head", + bn_name="_bn1", + model_name=self.name, + cur_stage=7) + self._pool = AdaptiveAvgPool2d(1) + + if self._global_params.dropout_rate: + self._drop = Dropout(p=self._global_params.dropout_rate, mode="upscale_in_train") + + param_attr, bias_attr = init_fc_layer("_fc") + self._fc = Linear(output_channels, class_dim, weight_attr=param_attr, bias_attr=bias_attr) + + if load_checkpoint is not None: + model_dict = paddle.load(load_checkpoint)[0] + self.set_dict(model_dict) + print("load custom checkpoint success") + + else: + checkpoint = os.path.join(self.directory, 'efficientnet_b4_imagenet.pdparams') + if not os.path.exists(checkpoint): + os.system( + 'wget https://paddlehub.bj.bcebos.com/dygraph/image_classification/efficientnet_b4_imagenet.pdparams -O ' + + checkpoint) + model_dict = paddle.load(checkpoint)[0] + self.set_dict(model_dict) + print("load pretrained checkpoint success") + + def forward(self, inputs: paddle.Tensor): + x = self._ef(inputs) + x = self._conv(x) + x = self._pool(x) + if self._global_params.dropout_rate: + x = self._drop(x) + x = paddle.squeeze(x, axis=[2, 3]) + x = self._fc(x) + return x diff --git a/hub_module/modules/image/classification/efficientnetb4_imagenet/processor.py b/hub_module/modules/image/classification/efficientnetb4_imagenet/processor.py deleted file mode 100644 index 12ccc9b2..00000000 --- a/hub_module/modules/image/classification/efficientnetb4_imagenet/processor.py +++ /dev/null @@ -1,69 +0,0 @@ -# -*- coding:utf-8 -*- -# copyright (c) 2020 PaddlePaddle Authors. All Rights Reserve. -# -# Licensed under the Apache License, Version 2.0 (the "License"); -# you may not use this file except in compliance with the License. -# You may obtain a copy of the License at -# -# http://www.apache.org/licenses/LICENSE-2.0 -# -# Unless required by applicable law or agreed to in writing, software -# distributed under the License is distributed on an "AS IS" BASIS, -# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. -# See the License for the specific language governing permissions and -# limitations under the License. - -from __future__ import absolute_import -from __future__ import division -from __future__ import print_function - -import base64 -import cv2 -import os - -import numpy as np - - -def base64_to_cv2(b64str): - data = base64.b64decode(b64str.encode('utf8')) - data = np.fromstring(data, np.uint8) - data = cv2.imdecode(data, cv2.IMREAD_COLOR) - return data - - -def softmax(x): - orig_shape = x.shape - if len(x.shape) > 1: - tmp = np.max(x, axis=1) - x -= tmp.reshape((x.shape[0], 1)) - x = np.exp(x) - tmp = np.sum(x, axis=1) - x /= tmp.reshape((x.shape[0], 1)) - else: - tmp = np.max(x) - x -= tmp - x = np.exp(x) - tmp = np.sum(x) - x /= tmp - return x - - -def postprocess(data_out, label_list, top_k): - """ - Postprocess output of network, one image at a time. - - Args: - data_out (numpy.ndarray): output data of network. - label_list (list): list of label. - top_k (int): Return top k results. - """ - output = [] - for result in data_out: - result_i = softmax(result) - output_i = {} - indexs = np.argsort(result_i)[::-1][0:top_k] - for index in indexs: - label = label_list[index].split(',')[0] - output_i[label] = float(result_i[index]) - output.append(output_i) - return output diff --git a/hub_module/modules/image/classification/efficientnetb5_imagenet/README.md b/hub_module/modules/image/classification/efficientnetb5_imagenet/README.md deleted file mode 100644 index 140e22c9..00000000 --- a/hub_module/modules/image/classification/efficientnetb5_imagenet/README.md +++ /dev/null @@ -1,149 +0,0 @@ -## 命令行预测 - -``` -hub run efficientnetb5_imagenet --input_path "/PATH/TO/IMAGE" -``` - -## API - -```python -def get_expected_image_width() -``` - -返回预处理的图片宽度，也就是224。 - -```python -def get_expected_image_height() -``` - -返回预处理的图片高度，也就是224。 - -```python -def get_pretrained_images_mean() -``` - -返回预处理的图片均值，也就是 \[0.485, 0.456, 0.406\]。 - -```python -def get_pretrained_images_std() -``` - -返回预处理的图片标准差，也就是 \[0.229, 0.224, 0.225\]。 - - -```python -def context(trainable=True, pretrained=True) -``` - -**参数** - -* trainable (bool): 计算图的参数是否为可训练的； -* pretrained (bool): 是否加载默认的预训练模型。 - -**返回** - -* inputs (dict): 计算图的输入，key 为 'image', value 为图片的张量； -* outputs (dict): 计算图的输出，key 为 'classification' 和 'feature_map'，其相应的值为： - * classification (paddle.fluid.framework.Variable): 分类结果，也就是全连接层的输出； - * feature\_map (paddle.fluid.framework.Variable): 特征匹配，全连接层前面的那个张量。 -* context\_prog(fluid.Program): 计算图，用于迁移学习。 - -```python -def classify(images=None, - paths=None, - batch_size=1, - use_gpu=False, - top_k=1): -``` - -**参数** - -* images (list\[numpy.ndarray\]): 图片数据，每一个图片数据的shape 均为 \[H, W, C\]，颜色空间为 BGR； -* paths (list\[str\]): 图片的路径； -* batch\_size (int): batch 的大小； -* use\_gpu (bool): 是否使用 GPU 来预测； -* top\_k (int): 返回预测结果的前 k 个。 - -**返回** - -res (list\[dict\]): 分类结果，列表的每一个元素均为字典，其中 key 为识别动物的类别，value为置信度。 - -```python -def save_inference_model(dirname, - model_filename=None, - params_filename=None, - combined=True) -``` - -将模型保存到指定路径。 - -**参数** - -* dirname: 存在模型的目录名称 -* model\_filename: 模型文件名称，默认为\_\_model\_\_ -* params\_filename: 参数文件名称，默认为\_\_params\_\_(仅当`combined`为True时生效) -* combined: 是否将参数保存到统一的一个文件中 - -## 代码示例 - -```python -import paddlehub as hub -import cv2 - -classifier = hub.Module(name="efficientnetb5_imagenet") - -result = classifier.classify(images=[cv2.imread('/PATH/TO/IMAGE')]) -# or -# result = classifier.classify(paths=['/PATH/TO/IMAGE']) -``` - -## 服务部署 - -PaddleHub Serving可以部署一个在线图像识别服务。 - -## 第一步：启动PaddleHub Serving - -运行启动命令： -```shell -$ hub serving start -m efficientnetb5_imagenet -``` - -这样就完成了一个在线图像识别服务化API的部署，默认端口号为8866。 - -**NOTE:** 如使用GPU预测，则需要在启动服务之前，请设置CUDA\_VISIBLE\_DEVICES环境变量，否则不用设置。 - -## 第二步：发送预测请求 - -配置好服务端，以下数行代码即可实现发送预测请求，获取预测结果 - -```python -import requests -import json -import cv2 -import base64 - - -def cv2_to_base64(image): - data = cv2.imencode('.jpg', image)[1] - return base64.b64encode(data.tostring()).decode('utf8') - - -# 发送HTTP请求 -data = {'images':[cv2_to_base64(cv2.imread("/PATH/TO/IMAGE"))]} -headers = {"Content-type": "application/json"} -url = "http://127.0.0.1:8866/predict/efficientnetb5_imagenet" -r = requests.post(url=url, headers=headers, data=json.dumps(data)) - -# 打印预测结果 -print(r.json()["results"]) -``` - -### 查看代码 - -https://github.com/PaddlePaddle/PaddleClas - -### 依赖 - -paddlepaddle >= 1.6.2 - -paddlehub >= 1.6.0 diff --git a/hub_module/modules/image/classification/efficientnetb5_imagenet/__init__.py b/hub_module/modules/image/classification/efficientnetb5_imagenet/__init__.py deleted file mode 100644 index 19a990da..00000000 --- a/hub_module/modules/image/classification/efficientnetb5_imagenet/__init__.py +++ /dev/null @@ -1 +0,0 @@ -# -*- coding:utf-8 -*- diff --git a/hub_module/modules/image/classification/efficientnetb5_imagenet/data_feed.py b/hub_module/modules/image/classification/efficientnetb5_imagenet/data_feed.py deleted file mode 100644 index 11181458..00000000 --- a/hub_module/modules/image/classification/efficientnetb5_imagenet/data_feed.py +++ /dev/null @@ -1,98 +0,0 @@ -# -*- coding:utf-8 -*- -# copyright (c) 2020 PaddlePaddle Authors. All Rights Reserve. -# -# Licensed under the Apache License, Version 2.0 (the "License"); -# you may not use this file except in compliance with the License. -# You may obtain a copy of the License at -# -# http://www.apache.org/licenses/LICENSE-2.0 -# -# Unless required by applicable law or agreed to in writing, software -# distributed under the License is distributed on an "AS IS" BASIS, -# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. -# See the License for the specific language governing permissions and -# limitations under the License. - -import os -import time -from collections import OrderedDict - -import cv2 -import numpy as np -from PIL import Image - -__all__ = ['reader'] - -DATA_DIM = 224 -img_mean = np.array([0.485, 0.456, 0.406]).reshape((3, 1, 1)) -img_std = np.array([0.229, 0.224, 0.225]).reshape((3, 1, 1)) - - -def resize_short(img, target_size): - percent = float(target_size) / min(img.size[0], img.size[1]) - resized_width = int(round(img.size[0] * percent)) - resized_height = int(round(img.size[1] * percent)) - img = img.resize((resized_width, resized_height), Image.LANCZOS) - return img - - -def crop_image(img, target_size, center): - width, height = img.size - size = target_size - if center == True: - w_start = (width - size) / 2 - h_start = (height - size) / 2 - else: - w_start = np.random.randint(0, width - size + 1) - h_start = np.random.randint(0, height - size + 1) - w_end = w_start + size - h_end = h_start + size - img = img.crop((w_start, h_start, w_end, h_end)) - return img - - -def process_image(img): - img = resize_short(img, target_size=256) - img = crop_image(img, target_size=DATA_DIM, center=True) - if img.mode != 'RGB': - img = img.convert('RGB') - img = np.array(img).astype('float32').transpose((2, 0, 1)) / 255 - img -= img_mean - img /= img_std - return img - - -def reader(images=None, paths=None): - """ - Preprocess to yield image. - - Args: - images (list[numpy.ndarray]): images data, shape of each is [H, W, C]. - paths (list[str]): paths to images. - - Yield: - each (collections.OrderedDict): info of original image, preprocessed image. - """ - component = list() - if paths: - for im_path in paths: - each = OrderedDict() - assert os.path.isfile( - im_path), "The {} isn't a valid file path.".format(im_path) - each['org_im_path'] = im_path - each['org_im'] = Image.open(im_path) - each['org_im_width'], each['org_im_height'] = each['org_im'].size - component.append(each) - if images is not None: - assert type(images), "images is a list." - for im in images: - each = OrderedDict() - each['org_im'] = Image.fromarray(im[:, :, ::-1]) - each['org_im_path'] = 'ndarray_time={}'.format( - round(time.time(), 6) * 1e6) - each['org_im_width'], each['org_im_height'] = each['org_im'].size - component.append(each) - - for element in component: - element['image'] = process_image(element['org_im']) - yield element diff --git a/hub_module/modules/image/classification/efficientnetb5_imagenet/efficientnet.py b/hub_module/modules/image/classification/efficientnetb5_imagenet/efficientnet.py deleted file mode 100644 index ef49e972..00000000 --- a/hub_module/modules/image/classification/efficientnetb5_imagenet/efficientnet.py +++ /dev/null @@ -1,623 +0,0 @@ -# -*- coding:utf-8 -*- -# copyright (c) 2020 PaddlePaddle Authors. All Rights Reserve. -# -# Licensed under the Apache License, Version 2.0 (the "License"); -# you may not use this file except in compliance with the License. -# You may obtain a copy of the License at -# -# http://www.apache.org/licenses/LICENSE-2.0 -# -# Unless required by applicable law or agreed to in writing, software -# distributed under the License is distributed on an "AS IS" BASIS, -# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. -# See the License for the specific language governing permissions and -# limitations under the License. - -from __future__ import absolute_import -from __future__ import division -from __future__ import print_function - -import collections -import re -import math -import copy - -import paddle.fluid as fluid -from efficientnetb5_imagenet.layers import conv2d, init_batch_norm_layer, init_fc_layer - -__all__ = [ - 'EfficientNet', 'EfficientNetB0_small', 'EfficientNetB0', 'EfficientNetB1', - 'EfficientNetB2', 'EfficientNetB3', 'EfficientNetB4', 'EfficientNetB5', - 'EfficientNetB6', 'EfficientNetB7' -] - -GlobalParams = collections.namedtuple('GlobalParams', [ - 'batch_norm_momentum', - 'batch_norm_epsilon', - 'dropout_rate', - 'num_classes', - 'width_coefficient', - 'depth_coefficient', - 'depth_divisor', - 'min_depth', - 'drop_connect_rate', -]) - -BlockArgs = collections.namedtuple('BlockArgs', [ - 'kernel_size', 'num_repeat', 'input_filters', 'output_filters', - 'expand_ratio', 'id_skip', 'stride', 'se_ratio' -]) - -GlobalParams.__new__.__defaults__ = (None, ) * len(GlobalParams._fields) -BlockArgs.__new__.__defaults__ = (None, ) * len(BlockArgs._fields) - - -def efficientnet_params(model_name): - """ Map EfficientNet model name to parameter coefficients. """ - params_dict = { - # Coefficients: width,depth,resolution,dropout - 'efficientnet-b0': (1.0, 1.0, 224, 0.2), - 'efficientnet-b1': (1.0, 1.1, 240, 0.2), - 'efficientnet-b2': (1.1, 1.2, 260, 0.3), - 'efficientnet-b3': (1.2, 1.4, 300, 0.3), - 'efficientnet-b4': (1.4, 1.8, 380, 0.4), - 'efficientnet-b5': (1.6, 2.2, 456, 0.4), - 'efficientnet-b6': (1.8, 2.6, 528, 0.5), - 'efficientnet-b7': (2.0, 3.1, 600, 0.5), - } - return params_dict[model_name] - - -def efficientnet(width_coefficient=None, - depth_coefficient=None, - dropout_rate=0.2, - drop_connect_rate=0.2): - """ Get block arguments according to parameter and coefficients. """ - blocks_args = [ - 'r1_k3_s11_e1_i32_o16_se0.25', - 'r2_k3_s22_e6_i16_o24_se0.25', - 'r2_k5_s22_e6_i24_o40_se0.25', - 'r3_k3_s22_e6_i40_o80_se0.25', - 'r3_k5_s11_e6_i80_o112_se0.25', - 'r4_k5_s22_e6_i112_o192_se0.25', - 'r1_k3_s11_e6_i192_o320_se0.25', - ] - blocks_args = BlockDecoder.decode(blocks_args) - - global_params = GlobalParams( - batch_norm_momentum=0.99, - batch_norm_epsilon=1e-3, - dropout_rate=dropout_rate, - drop_connect_rate=drop_connect_rate, - num_classes=1000, - width_coefficient=width_coefficient, - depth_coefficient=depth_coefficient, - depth_divisor=8, - min_depth=None) - - return blocks_args, global_params - - -def get_model_params(model_name, override_params): - """ Get the block args and global params for a given model """ - if model_name.startswith('efficientnet'): - w, d, _, p = efficientnet_params(model_name) - blocks_args, global_params = efficientnet( - width_coefficient=w, depth_coefficient=d, dropout_rate=p) - else: - raise NotImplementedError( - 'model name is not pre-defined: %s' % model_name) - if override_params: - global_params = global_params._replace(**override_params) - return blocks_args, global_params - - -def round_filters(filters, global_params): - """ Calculate and round number of filters based on depth multiplier. """ - multiplier = global_params.width_coefficient - if not multiplier: - return filters - divisor = global_params.depth_divisor - min_depth = global_params.min_depth - filters *= multiplier - min_depth = min_depth or divisor - new_filters = max(min_depth, - int(filters + divisor / 2) // divisor * divisor) - if new_filters < 0.9 * filters: # prevent rounding by more than 10% - new_filters += divisor - return int(new_filters) - - -def round_repeats(repeats, global_params): - """ Round number of filters based on depth multiplier. """ - multiplier = global_params.depth_coefficient - if not multiplier: - return repeats - return int(math.ceil(multiplier * repeats)) - - -class EfficientNet(): - def __init__(self, - name='b0', - padding_type='SAME', - override_params=None, - is_test=False, - use_se=True): - valid_names = ['b' + str(i) for i in range(8)] - assert name in valid_names, 'efficient name should be in b0~b7' - model_name = 'efficientnet-' + name - self._blocks_args, self._global_params = get_model_params( - model_name, override_params) - self._bn_mom = self._global_params.batch_norm_momentum - self._bn_eps = self._global_params.batch_norm_epsilon - self.padding_type = padding_type - self.use_se = use_se - - def net(self, input, class_dim=1000, is_test=False): - - conv = self.extract_features(input, is_test=is_test) - - out_channels = round_filters(1280, self._global_params) - conv = self.conv_bn_layer( - conv, - num_filters=out_channels, - filter_size=1, - bn_act='swish', - bn_mom=self._bn_mom, - bn_eps=self._bn_eps, - padding_type=self.padding_type, - name='', - conv_name='_conv_head', - bn_name='_bn1') - - pool = fluid.layers.pool2d( - input=conv, pool_type='avg', global_pooling=True, use_cudnn=False) - - if not is_test and self._global_params.dropout_rate: - pool = fluid.layers.dropout( - pool, - self._global_params.dropout_rate, - dropout_implementation='upscale_in_train') - - param_attr, bias_attr = init_fc_layer(class_dim, '_fc') - out = fluid.layers.fc( - pool, - class_dim, - name='_fc', - param_attr=param_attr, - bias_attr=bias_attr) - return out, pool - - def _drop_connect(self, inputs, prob, is_test): - if is_test: - return inputs - keep_prob = 1.0 - prob - random_tensor = keep_prob + fluid.layers.uniform_random_batch_size_like( - inputs, [-1, 1, 1, 1], min=0., max=1.) - binary_tensor = fluid.layers.floor(random_tensor) - output = inputs / keep_prob * binary_tensor - return output - - def _expand_conv_norm(self, inputs, block_args, is_test, name=None): - # Expansion phase - oup = block_args.input_filters * block_args.expand_ratio # number of output channels - - if block_args.expand_ratio != 1: - conv = self.conv_bn_layer( - inputs, - num_filters=oup, - filter_size=1, - bn_act=None, - bn_mom=self._bn_mom, - bn_eps=self._bn_eps, - padding_type=self.padding_type, - name=name, - conv_name=name + '_expand_conv', - bn_name='_bn0') - - return conv - - def _depthwise_conv_norm(self, inputs, block_args, is_test, name=None): - k = block_args.kernel_size - s = block_args.stride - if isinstance(s, list) or isinstance(s, tuple): - s = s[0] - oup = block_args.input_filters * block_args.expand_ratio # number of output channels - - conv = self.conv_bn_layer( - inputs, - num_filters=oup, - filter_size=k, - stride=s, - num_groups=oup, - bn_act=None, - padding_type=self.padding_type, - bn_mom=self._bn_mom, - bn_eps=self._bn_eps, - name=name, - use_cudnn=False, - conv_name=name + '_depthwise_conv', - bn_name='_bn1') - - return conv - - def _project_conv_norm(self, inputs, block_args, is_test, name=None): - final_oup = block_args.output_filters - conv = self.conv_bn_layer( - inputs, - num_filters=final_oup, - filter_size=1, - bn_act=None, - padding_type=self.padding_type, - bn_mom=self._bn_mom, - bn_eps=self._bn_eps, - name=name, - conv_name=name + '_project_conv', - bn_name='_bn2') - return conv - - def conv_bn_layer(self, - input, - filter_size, - num_filters, - stride=1, - num_groups=1, - padding_type="SAME", - conv_act=None, - bn_act='swish', - use_cudnn=True, - use_bn=True, - bn_mom=0.9, - bn_eps=1e-05, - use_bias=False, - name=None, - conv_name=None, - bn_name=None): - conv = conv2d( - input=input, - num_filters=num_filters, - filter_size=filter_size, - stride=stride, - groups=num_groups, - act=conv_act, - padding_type=padding_type, - use_cudnn=use_cudnn, - name=conv_name, - use_bias=use_bias) - - if use_bn == False: - return conv - else: - bn_name = name + bn_name - param_attr, bias_attr = init_batch_norm_layer(bn_name) - return fluid.layers.batch_norm( - input=conv, - act=bn_act, - momentum=bn_mom, - epsilon=bn_eps, - name=bn_name, - moving_mean_name=bn_name + '_mean', - moving_variance_name=bn_name + '_variance', - param_attr=param_attr, - bias_attr=bias_attr) - - def _conv_stem_norm(self, inputs, is_test): - out_channels = round_filters(32, self._global_params) - bn = self.conv_bn_layer( - inputs, - num_filters=out_channels, - filter_size=3, - stride=2, - bn_act=None, - bn_mom=self._bn_mom, - padding_type=self.padding_type, - bn_eps=self._bn_eps, - name='', - conv_name='_conv_stem', - bn_name='_bn0') - - return bn - - def mb_conv_block(self, - inputs, - block_args, - is_test=False, - drop_connect_rate=None, - name=None): - # Expansion and Depthwise Convolution - oup = block_args.input_filters * block_args.expand_ratio # number of output channels - has_se = self.use_se and (block_args.se_ratio is - not None) and (0 < block_args.se_ratio <= 1) - id_skip = block_args.id_skip # skip connection and drop connect - conv = inputs - if block_args.expand_ratio != 1: - conv = fluid.layers.swish( - self._expand_conv_norm(conv, block_args, is_test, name)) - - conv = fluid.layers.swish( - self._depthwise_conv_norm(conv, block_args, is_test, name)) - - # Squeeze and Excitation - if has_se: - num_squeezed_channels = max( - 1, int(block_args.input_filters * block_args.se_ratio)) - conv = self.se_block(conv, num_squeezed_channels, oup, name) - - conv = self._project_conv_norm(conv, block_args, is_test, name) - - # Skip connection and drop connect - input_filters, output_filters = block_args.input_filters, block_args.output_filters - if id_skip and block_args.stride == 1 and input_filters == output_filters: - if drop_connect_rate: - conv = self._drop_connect(conv, drop_connect_rate, is_test) - conv = fluid.layers.elementwise_add(conv, inputs) - - return conv - - def se_block(self, inputs, num_squeezed_channels, oup, name): - x_squeezed = fluid.layers.pool2d( - input=inputs, pool_type='avg', global_pooling=True, use_cudnn=False) - x_squeezed = conv2d( - x_squeezed, - num_filters=num_squeezed_channels, - filter_size=1, - use_bias=True, - padding_type=self.padding_type, - act='swish', - name=name + '_se_reduce') - x_squeezed = conv2d( - x_squeezed, - num_filters=oup, - filter_size=1, - use_bias=True, - padding_type=self.padding_type, - name=name + '_se_expand') - se_out = inputs * fluid.layers.sigmoid(x_squeezed) - return se_out - - def extract_features(self, inputs, is_test): - """ Returns output of the final convolution layer """ - - conv = fluid.layers.swish(self._conv_stem_norm(inputs, is_test=is_test)) - - block_args_copy = copy.deepcopy(self._blocks_args) - idx = 0 - block_size = 0 - for block_arg in block_args_copy: - block_arg = block_arg._replace( - input_filters=round_filters(block_arg.input_filters, - self._global_params), - output_filters=round_filters(block_arg.output_filters, - self._global_params), - num_repeat=round_repeats(block_arg.num_repeat, - self._global_params)) - block_size += 1 - for _ in range(block_arg.num_repeat - 1): - block_size += 1 - - for block_args in self._blocks_args: - - # Update block input and output filters based on depth multiplier. - block_args = block_args._replace( - input_filters=round_filters(block_args.input_filters, - self._global_params), - output_filters=round_filters(block_args.output_filters, - self._global_params), - num_repeat=round_repeats(block_args.num_repeat, - self._global_params)) - - # The first block needs to take care of stride and filter size increase. - drop_connect_rate = self._global_params.drop_connect_rate - if drop_connect_rate: - drop_connect_rate *= float(idx) / block_size - conv = self.mb_conv_block(conv, block_args, is_test, - drop_connect_rate, - '_blocks.' + str(idx) + '.') - - idx += 1 - if block_args.num_repeat > 1: - block_args = block_args._replace( - input_filters=block_args.output_filters, stride=1) - for _ in range(block_args.num_repeat - 1): - drop_connect_rate = self._global_params.drop_connect_rate - if drop_connect_rate: - drop_connect_rate *= float(idx) / block_size - conv = self.mb_conv_block(conv, block_args, is_test, - drop_connect_rate, - '_blocks.' + str(idx) + '.') - idx += 1 - - return conv - - def shortcut(self, input, data_residual): - return fluid.layers.elementwise_add(input, data_residual) - - -class BlockDecoder(object): - """ Block Decoder for readability, straight from the official TensorFlow repository """ - - @staticmethod - def _decode_block_string(block_string): - """ Gets a block through a string notation of arguments. """ - assert isinstance(block_string, str) - - ops = block_string.split('_') - options = {} - for op in ops: - splits = re.split(r'(\d.*)', op) - if len(splits) >= 2: - key, value = splits[:2] - options[key] = value - - # Check stride - assert (('s' in options and len(options['s']) == 1) or - (len(options['s']) == 2 and options['s'][0] == options['s'][1])) - - return BlockArgs( - kernel_size=int(options['k']), - num_repeat=int(options['r']), - input_filters=int(options['i']), - output_filters=int(options['o']), - expand_ratio=int(options['e']), - id_skip=('noskip' not in block_string), - se_ratio=float(options['se']) if 'se' in options else None, - stride=[int(options['s'][0])]) - - @staticmethod - def _encode_block_string(block): - """Encodes a block to a string.""" - args = [ - 'r%d' % block.num_repeat, - 'k%d' % block.kernel_size, - 's%d%d' % (block.strides[0], block.strides[1]), - 'e%s' % block.expand_ratio, - 'i%d' % block.input_filters, - 'o%d' % block.output_filters - ] - if 0 < block.se_ratio <= 1: - args.append('se%s' % block.se_ratio) - if block.id_skip is False: - args.append('noskip') - return '_'.join(args) - - @staticmethod - def decode(string_list): - """ - Decodes a list of string notations to specify blocks inside the network. - :param string_list: a list of strings, each string is a notation of block - :return: a list of BlockArgs namedtuples of block args - """ - assert isinstance(string_list, list) - blocks_args = [] - for block_string in string_list: - blocks_args.append(BlockDecoder._decode_block_string(block_string)) - return blocks_args - - @staticmethod - def encode(blocks_args): - """ - Encodes a list of BlockArgs to a list of strings. - :param blocks_args: a list of BlockArgs namedtuples of block args - :return: a list of strings, each string is a notation of block - """ - block_strings = [] - for block in blocks_args: - block_strings.append(BlockDecoder._encode_block_string(block)) - return block_strings - - -def EfficientNetB0_small(is_test=False, - padding_type='SAME', - override_params=None, - use_se=False): - model = EfficientNet( - name='b0', - is_test=is_test, - padding_type=padding_type, - override_params=override_params, - use_se=use_se) - return model - - -def EfficientNetB0(is_test=False, - padding_type='SAME', - override_params=None, - use_se=True): - model = EfficientNet( - name='b0', - is_test=is_test, - padding_type=padding_type, - override_params=override_params, - use_se=use_se) - return model - - -def EfficientNetB1(is_test=False, - padding_type='SAME', - override_params=None, - use_se=True): - model = EfficientNet( - name='b1', - is_test=is_test, - padding_type=padding_type, - override_params=override_params, - use_se=use_se) - return model - - -def EfficientNetB2(is_test=False, - padding_type='SAME', - override_params=None, - use_se=True): - model = EfficientNet( - name='b2', - is_test=is_test, - padding_type=padding_type, - override_params=override_params, - use_se=use_se) - return model - - -def EfficientNetB3(is_test=False, - padding_type='SAME', - override_params=None, - use_se=True): - model = EfficientNet( - name='b3', - is_test=is_test, - padding_type=padding_type, - override_params=override_params, - use_se=use_se) - return model - - -def EfficientNetB4(is_test=False, - padding_type='SAME', - override_params=None, - use_se=True): - model = EfficientNet( - name='b4', - is_test=is_test, - padding_type=padding_type, - override_params=override_params, - use_se=use_se) - return model - - -def EfficientNetB5(is_test=False, - padding_type='SAME', - override_params=None, - use_se=True): - model = EfficientNet( - name='b5', - is_test=is_test, - padding_type=padding_type, - override_params=override_params, - use_se=use_se) - return model - - -def EfficientNetB6(is_test=False, - padding_type='SAME', - override_params=None, - use_se=True): - model = EfficientNet( - name='b6', - is_test=is_test, - padding_type=padding_type, - override_params=override_params, - use_se=use_se) - return model - - -def EfficientNetB7(is_test=False, - padding_type='SAME', - override_params=None, - use_se=True): - model = EfficientNet( - name='b7', - is_test=is_test, - padding_type=padding_type, - override_params=override_params, - use_se=use_se) - return model diff --git a/hub_module/modules/image/classification/efficientnetb5_imagenet/label_list.txt b/hub_module/modules/image/classification/efficientnetb5_imagenet/label_list.txt deleted file mode 100644 index a509c007..00000000 --- a/hub_module/modules/image/classification/efficientnetb5_imagenet/label_list.txt +++ /dev/null @@ -1,1000 +0,0 @@ -tench, Tinca tinca -goldfish, Carassius auratus -great white shark, white shark, man-eater, man-eating shark, Carcharodon carcharias -tiger shark, Galeocerdo cuvieri -hammerhead, hammerhead shark -electric ray, crampfish, numbfish, torpedo -stingray -cock -hen -ostrich, Struthio camelus -brambling, Fringilla montifringilla -goldfinch, Carduelis carduelis -house finch, linnet, Carpodacus mexicanus -junco, snowbird -indigo bunting, indigo finch, indigo bird, Passerina cyanea -robin, American robin, Turdus migratorius -bulbul -jay -magpie -chickadee -water ouzel, dipper -kite -bald eagle, American eagle, Haliaeetus leucocephalus -vulture -great grey owl, great gray owl, Strix nebulosa -European fire salamander, Salamandra salamandra -common newt, Triturus vulgaris -eft -spotted salamander, Ambystoma maculatum -axolotl, mud puppy, Ambystoma mexicanum -bullfrog, Rana catesbeiana -tree frog, tree-frog -tailed frog, bell toad, ribbed toad, tailed toad, Ascaphus trui -loggerhead, loggerhead turtle, Caretta caretta -leatherback turtle, leatherback, leathery turtle, Dermochelys coriacea -mud turtle -terrapin -box turtle, box tortoise -banded gecko -common iguana, iguana, Iguana iguana -American chameleon, anole, Anolis carolinensis -whiptail, whiptail lizard -agama -frilled lizard, Chlamydosaurus kingi -alligator lizard -Gila monster, Heloderma suspectum -green lizard, Lacerta viridis -African chameleon, Chamaeleo chamaeleon -Komodo dragon, Komodo lizard, dragon lizard, giant lizard, Varanus komodoensis -African crocodile, Nile crocodile, Crocodylus niloticus -American alligator, Alligator mississipiensis -triceratops -thunder snake, worm snake, Carphophis amoenus -ringneck snake, ring-necked snake, ring snake -hognose snake, puff adder, sand viper -green snake, grass snake -king snake, kingsnake -garter snake, grass snake -water snake -vine snake -night snake, Hypsiglena torquata -boa constrictor, Constrictor constrictor -rock python, rock snake, Python sebae -Indian cobra, Naja naja -green mamba -sea snake -horned viper, cerastes, sand viper, horned asp, Cerastes cornutus -diamondback, diamondback rattlesnake, Crotalus adamanteus -sidewinder, horned rattlesnake, Crotalus cerastes -trilobite -harvestman, daddy longlegs, Phalangium opilio -scorpion -black and gold garden spider, Argiope aurantia -barn spider, Araneus cavaticus -garden spider, Aranea diademata -black widow, Latrodectus mactans -tarantula -wolf spider, hunting spider -tick -centipede -black grouse -ptarmigan -ruffed grouse, partridge, Bonasa umbellus -prairie chicken, prairie grouse, prairie fowl -peacock -quail -partridge -African grey, African gray, Psittacus erithacus -macaw -sulphur-crested cockatoo, Kakatoe galerita, Cacatua galerita -lorikeet -coucal -bee eater -hornbill -hummingbird -jacamar -toucan -drake -red-breasted merganser, Mergus serrator -goose -black swan, Cygnus atratus -tusker -echidna, spiny anteater, anteater -platypus, duckbill, duckbilled platypus, duck-billed platypus, Ornithorhynchus anatinus -wallaby, brush kangaroo -koala, koala bear, kangaroo bear, native bear, Phascolarctos cinereus -wombat -jellyfish -sea anemone, anemone -brain coral -flatworm, platyhelminth -nematode, nematode worm, roundworm -conch -snail -slug -sea slug, nudibranch -chiton, coat-of-mail shell, sea cradle, polyplacophore -chambered nautilus, pearly nautilus, nautilus -Dungeness crab, Cancer magister -rock crab, Cancer irroratus -fiddler crab -king crab, Alaska crab, Alaskan king crab, Alaska king crab, Paralithodes camtschatica -American lobster, Northern lobster, Maine lobster, Homarus americanus -spiny lobster, langouste, rock lobster, crawfish, crayfish, sea crawfish -crayfish, crawfish, crawdad, crawdaddy -hermit crab -isopod -white stork, Ciconia ciconia -black stork, Ciconia nigra -spoonbill -flamingo -little blue heron, Egretta caerulea -American egret, great white heron, Egretta albus -bittern -crane -limpkin, Aramus pictus -European gallinule, Porphyrio porphyrio -American coot, marsh hen, mud hen, water hen, Fulica americana -bustard -ruddy turnstone, Arenaria interpres -red-backed sandpiper, dunlin, Erolia alpina -redshank, Tringa totanus -dowitcher -oystercatcher, oyster catcher -pelican -king penguin, Aptenodytes patagonica -albatross, mollymawk -grey whale, gray whale, devilfish, Eschrichtius gibbosus, Eschrichtius robustus -killer whale, killer, orca, grampus, sea wolf, Orcinus orca -dugong, Dugong dugon -sea lion -Chihuahua -Japanese spaniel -Maltese dog, Maltese terrier, Maltese -Pekinese, Pekingese, Peke -Shih-Tzu -Blenheim spaniel -papillon -toy terrier -Rhodesian ridgeback -Afghan hound, Afghan -basset, basset hound -beagle -bloodhound, sleuthhound -bluetick -black-and-tan coonhound -Walker hound, Walker foxhound -English foxhound -redbone -borzoi, Russian wolfhound -Irish wolfhound -Italian greyhound -whippet -Ibizan hound, Ibizan Podenco -Norwegian elkhound, elkhound -otterhound, otter hound -Saluki, gazelle hound -Scottish deerhound, deerhound -Weimaraner -Staffordshire bullterrier, Staffordshire bull terrier -American Staffordshire terrier, Staffordshire terrier, American pit bull terrier, pit bull terrier -Bedlington terrier -Border terrier -Kerry blue terrier -Irish terrier -Norfolk terrier -Norwich terrier -Yorkshire terrier -wire-haired fox terrier -Lakeland terrier -Sealyham terrier, Sealyham -Airedale, Airedale terrier -cairn, cairn terrier -Australian terrier -Dandie Dinmont, Dandie Dinmont terrier -Boston bull, Boston terrier -miniature schnauzer -giant schnauzer -standard schnauzer -Scotch terrier, Scottish terrier, Scottie -Tibetan terrier, chrysanthemum dog -silky terrier, Sydney silky -soft-coated wheaten terrier -West Highland white terrier -Lhasa, Lhasa apso -flat-coated retriever -curly-coated retriever -golden retriever -Labrador retriever -Chesapeake Bay retriever -German short-haired pointer -vizsla, Hungarian pointer -English setter -Irish setter, red setter -Gordon setter -Brittany spaniel -clumber, clumber spaniel -English springer, English springer spaniel -Welsh springer spaniel -cocker spaniel, English cocker spaniel, cocker -Sussex spaniel -Irish water spaniel -kuvasz -schipperke -groenendael -malinois -briard -kelpie -komondor -Old English sheepdog, bobtail -Shetland sheepdog, Shetland sheep dog, Shetland -collie -Border collie -Bouvier des Flandres, Bouviers des Flandres -Rottweiler -German shepherd, German shepherd dog, German police dog, alsatian -Doberman, Doberman pinscher -miniature pinscher -Greater Swiss Mountain dog -Bernese mountain dog -Appenzeller -EntleBucher -boxer -bull mastiff -Tibetan mastiff -French bulldog -Great Dane -Saint Bernard, St Bernard -Eskimo dog, husky -malamute, malemute, Alaskan malamute -Siberian husky -dalmatian, coach dog, carriage dog -affenpinscher, monkey pinscher, monkey dog -basenji -pug, pug-dog -Leonberg -Newfoundland, Newfoundland dog -Great Pyrenees -Samoyed, Samoyede -Pomeranian -chow, chow chow -keeshond -Brabancon griffon -Pembroke, Pembroke Welsh corgi -Cardigan, Cardigan Welsh corgi -toy poodle -miniature poodle -standard poodle -Mexican hairless -timber wolf, grey wolf, gray wolf, Canis lupus -white wolf, Arctic wolf, Canis lupus tundrarum -red wolf, maned wolf, Canis rufus, Canis niger -coyote, prairie wolf, brush wolf, Canis latrans -dingo, warrigal, warragal, Canis dingo -dhole, Cuon alpinus -African hunting dog, hyena dog, Cape hunting dog, Lycaon pictus -hyena, hyaena -red fox, Vulpes vulpes -kit fox, Vulpes macrotis -Arctic fox, white fox, Alopex lagopus -grey fox, gray fox, Urocyon cinereoargenteus -tabby, tabby cat -tiger cat -Persian cat -Siamese cat, Siamese -Egyptian cat -cougar, puma, catamount, mountain lion, painter, panther, Felis concolor -lynx, catamount -leopard, Panthera pardus -snow leopard, ounce, Panthera uncia -jaguar, panther, Panthera onca, Felis onca -lion, king of beasts, Panthera leo -tiger, Panthera tigris -cheetah, chetah, Acinonyx jubatus -brown bear, bruin, Ursus arctos -American black bear, black bear, Ursus americanus, Euarctos americanus -ice bear, polar bear, Ursus Maritimus, Thalarctos maritimus -sloth bear, Melursus ursinus, Ursus ursinus -mongoose -meerkat, mierkat -tiger beetle -ladybug, ladybeetle, lady beetle, ladybird, ladybird beetle -ground beetle, carabid beetle -long-horned beetle, longicorn, longicorn beetle -leaf beetle, chrysomelid -dung beetle -rhinoceros beetle -weevil -fly -bee -ant, emmet, pismire -grasshopper, hopper -cricket -walking stick, walkingstick, stick insect -cockroach, roach -mantis, mantid -cicada, cicala -leafhopper -lacewing, lacewing fly -dragonfly, darning needle, devil's darning needle, sewing needle, snake feeder, snake doctor, mosquito hawk, skeeter hawk -damselfly -admiral -ringlet, ringlet butterfly -monarch, monarch butterfly, milkweed butterfly, Danaus plexippus -cabbage butterfly -sulphur butterfly, sulfur butterfly -lycaenid, lycaenid butterfly -starfish, sea star -sea urchin -sea cucumber, holothurian -wood rabbit, cottontail, cottontail rabbit -hare -Angora, Angora rabbit -hamster -porcupine, hedgehog -fox squirrel, eastern fox squirrel, Sciurus niger -marmot -beaver -guinea pig, Cavia cobaya -sorrel -zebra -hog, pig, grunter, squealer, Sus scrofa -wild boar, boar, Sus scrofa -warthog -hippopotamus, hippo, river horse, Hippopotamus amphibius -ox -water buffalo, water ox, Asiatic buffalo, Bubalus bubalis -bison -ram, tup -bighorn, bighorn sheep, cimarron, Rocky Mountain bighorn, Rocky Mountain sheep, Ovis canadensis -ibex, Capra ibex -hartebeest -impala, Aepyceros melampus -gazelle -Arabian camel, dromedary, Camelus dromedarius -llama -weasel -mink -polecat, fitch, foulmart, foumart, Mustela putorius -black-footed ferret, ferret, Mustela nigripes -otter -skunk, polecat, wood pussy -badger -armadillo -three-toed sloth, ai, Bradypus tridactylus -orangutan, orang, orangutang, Pongo pygmaeus -gorilla, Gorilla gorilla -chimpanzee, chimp, Pan troglodytes -gibbon, Hylobates lar -siamang, Hylobates syndactylus, Symphalangus syndactylus -guenon, guenon monkey -patas, hussar monkey, Erythrocebus patas -baboon -macaque -langur -colobus, colobus monkey -proboscis monkey, Nasalis larvatus -marmoset -capuchin, ringtail, Cebus capucinus -howler monkey, howler -titi, titi monkey -spider monkey, Ateles geoffroyi -squirrel monkey, Saimiri sciureus -Madagascar cat, ring-tailed lemur, Lemur catta -indri, indris, Indri indri, Indri brevicaudatus -Indian elephant, Elephas maximus -African elephant, Loxodonta africana -lesser panda, red panda, panda, bear cat, cat bear, Ailurus fulgens -giant panda, panda, panda bear, coon bear, Ailuropoda melanoleuca -barracouta, snoek -eel -coho, cohoe, coho salmon, blue jack, silver salmon, Oncorhynchus kisutch -rock beauty, Holocanthus tricolor -anemone fish -sturgeon -gar, garfish, garpike, billfish, Lepisosteus osseus -lionfish -puffer, pufferfish, blowfish, globefish -abacus -abaya -academic gown, academic robe, judge's robe -accordion, piano accordion, squeeze box -acoustic guitar -aircraft carrier, carrier, flattop, attack aircraft carrier -airliner -airship, dirigible -altar -ambulance -amphibian, amphibious vehicle -analog clock -apiary, bee house -apron -ashcan, trash can, garbage can, wastebin, ash bin, ash-bin, ashbin, dustbin, trash barrel, trash bin -assault rifle, assault gun -backpack, back pack, knapsack, packsack, rucksack, haversack -bakery, bakeshop, bakehouse -balance beam, beam -balloon -ballpoint, ballpoint pen, ballpen, Biro -Band Aid -banjo -bannister, banister, balustrade, balusters, handrail -barbell -barber chair -barbershop -barn -barometer -barrel, cask -barrow, garden cart, lawn cart, wheelbarrow -baseball -basketball -bassinet -bassoon -bathing cap, swimming cap -bath towel -bathtub, bathing tub, bath, tub -beach wagon, station wagon, wagon, estate car, beach waggon, station waggon, waggon -beacon, lighthouse, beacon light, pharos -beaker -bearskin, busby, shako -beer bottle -beer glass -bell cote, bell cot -bib -bicycle-built-for-two, tandem bicycle, tandem -bikini, two-piece -binder, ring-binder -binoculars, field glasses, opera glasses -birdhouse -boathouse -bobsled, bobsleigh, bob -bolo tie, bolo, bola tie, bola -bonnet, poke bonnet -bookcase -bookshop, bookstore, bookstall -bottlecap -bow -bow tie, bow-tie, bowtie -brass, memorial tablet, plaque -brassiere, bra, bandeau -breakwater, groin, groyne, mole, bulwark, seawall, jetty -breastplate, aegis, egis -broom -bucket, pail -buckle -bulletproof vest -bullet train, bullet -butcher shop, meat market -cab, hack, taxi, taxicab -caldron, cauldron -candle, taper, wax light -cannon -canoe -can opener, tin opener -cardigan -car mirror -carousel, carrousel, merry-go-round, roundabout, whirligig -carpenter's kit, tool kit -carton -car wheel -cash machine, cash dispenser, automated teller machine, automatic teller machine, automated teller, automatic teller, ATM -cassette -cassette player -castle -catamaran -CD player -cello, violoncello -cellular telephone, cellular phone, cellphone, cell, mobile phone -chain -chainlink fence -chain mail, ring mail, mail, chain armor, chain armour, ring armor, ring armour -chain saw, chainsaw -chest -chiffonier, commode -chime, bell, gong -china cabinet, china closet -Christmas stocking -church, church building -cinema, movie theater, movie theatre, movie house, picture palace -cleaver, meat cleaver, chopper -cliff dwelling -cloak -clog, geta, patten, sabot -cocktail shaker -coffee mug -coffeepot -coil, spiral, volute, whorl, helix -combination lock -computer keyboard, keypad -confectionery, confectionary, candy store -container ship, containership, container vessel -convertible -corkscrew, bottle screw -cornet, horn, trumpet, trump -cowboy boot -cowboy hat, ten-gallon hat -cradle -crane -crash helmet -crate -crib, cot -Crock Pot -croquet ball -crutch -cuirass -dam, dike, dyke -desk -desktop computer -dial telephone, dial phone -diaper, nappy, napkin -digital clock -digital watch -dining table, board -dishrag, dishcloth -dishwasher, dish washer, dishwashing machine -disk brake, disc brake -dock, dockage, docking facility -dogsled, dog sled, dog sleigh -dome -doormat, welcome mat -drilling platform, offshore rig -drum, membranophone, tympan -drumstick -dumbbell -Dutch oven -electric fan, blower -electric guitar -electric locomotive -entertainment center -envelope -espresso maker -face powder -feather boa, boa -file, file cabinet, filing cabinet -fireboat -fire engine, fire truck -fire screen, fireguard -flagpole, flagstaff -flute, transverse flute -folding chair -football helmet -forklift -fountain -fountain pen -four-poster -freight car -French horn, horn -frying pan, frypan, skillet -fur coat -garbage truck, dustcart -gasmask, respirator, gas helmet -gas pump, gasoline pump, petrol pump, island dispenser -goblet -go-kart -golf ball -golfcart, golf cart -gondola -gong, tam-tam -gown -grand piano, grand -greenhouse, nursery, glasshouse -grille, radiator grille -grocery store, grocery, food market, market -guillotine -hair slide -hair spray -half track -hammer -hamper -hand blower, blow dryer, blow drier, hair dryer, hair drier -hand-held computer, hand-held microcomputer -handkerchief, hankie, hanky, hankey -hard disc, hard disk, fixed disk -harmonica, mouth organ, harp, mouth harp -harp -harvester, reaper -hatchet -holster -home theater, home theatre -honeycomb -hook, claw -hoopskirt, crinoline -horizontal bar, high bar -horse cart, horse-cart -hourglass -iPod -iron, smoothing iron -jack-o'-lantern -jean, blue jean, denim -jeep, landrover -jersey, T-shirt, tee shirt -jigsaw puzzle -jinrikisha, ricksha, rickshaw -joystick -kimono -knee pad -knot -lab coat, laboratory coat -ladle -lampshade, lamp shade -laptop, laptop computer -lawn mower, mower -lens cap, lens cover -letter opener, paper knife, paperknife -library -lifeboat -lighter, light, igniter, ignitor -limousine, limo -liner, ocean liner -lipstick, lip rouge -Loafer -lotion -loudspeaker, speaker, speaker unit, loudspeaker system, speaker system -loupe, jeweler's loupe -lumbermill, sawmill -magnetic compass -mailbag, postbag -mailbox, letter box -maillot -maillot, tank suit -manhole cover -maraca -marimba, xylophone -mask -matchstick -maypole -maze, labyrinth -measuring cup -medicine chest, medicine cabinet -megalith, megalithic structure -microphone, mike -microwave, microwave oven -military uniform -milk can -minibus -miniskirt, mini -minivan -missile -mitten -mixing bowl -mobile home, manufactured home -Model T -modem -monastery -monitor -moped -mortar -mortarboard -mosque -mosquito net -motor scooter, scooter -mountain bike, all-terrain bike, off-roader -mountain tent -mouse, computer mouse -mousetrap -moving van -muzzle -nail -neck brace -necklace -nipple -notebook, notebook computer -obelisk -oboe, hautboy, hautbois -ocarina, sweet potato -odometer, hodometer, mileometer, milometer -oil filter -organ, pipe organ -oscilloscope, scope, cathode-ray oscilloscope, CRO -overskirt -oxcart -oxygen mask -packet -paddle, boat paddle -paddlewheel, paddle wheel -padlock -paintbrush -pajama, pyjama, pj's, jammies -palace -panpipe, pandean pipe, syrinx -paper towel -parachute, chute -parallel bars, bars -park bench -parking meter -passenger car, coach, carriage -patio, terrace -pay-phone, pay-station -pedestal, plinth, footstall -pencil box, pencil case -pencil sharpener -perfume, essence -Petri dish -photocopier -pick, plectrum, plectron -pickelhaube -picket fence, paling -pickup, pickup truck -pier -piggy bank, penny bank -pill bottle -pillow -ping-pong ball -pinwheel -pirate, pirate ship -pitcher, ewer -plane, carpenter's plane, woodworking plane -planetarium -plastic bag -plate rack -plow, plough -plunger, plumber's helper -Polaroid camera, Polaroid Land camera -pole -police van, police wagon, paddy wagon, patrol wagon, wagon, black Maria -poncho -pool table, billiard table, snooker table -pop bottle, soda bottle -pot, flowerpot -potter's wheel -power drill -prayer rug, prayer mat -printer -prison, prison house -projectile, missile -projector -puck, hockey puck -punching bag, punch bag, punching ball, punchball -purse -quill, quill pen -quilt, comforter, comfort, puff -racer, race car, racing car -racket, racquet -radiator -radio, wireless -radio telescope, radio reflector -rain barrel -recreational vehicle, RV, R.V. -reel -reflex camera -refrigerator, icebox -remote control, remote -restaurant, eating house, eating place, eatery -revolver, six-gun, six-shooter -rifle -rocking chair, rocker -rotisserie -rubber eraser, rubber, pencil eraser -rugby ball -rule, ruler -running shoe -safe -safety pin -saltshaker, salt shaker -sandal -sarong -sax, saxophone -scabbard -scale, weighing machine -school bus -schooner -scoreboard -screen, CRT screen -screw -screwdriver -seat belt, seatbelt -sewing machine -shield, buckler -shoe shop, shoe-shop, shoe store -shoji -shopping basket -shopping cart -shovel -shower cap -shower curtain -ski -ski mask -sleeping bag -slide rule, slipstick -sliding door -slot, one-armed bandit -snorkel -snowmobile -snowplow, snowplough -soap dispenser -soccer ball -sock -solar dish, solar collector, solar furnace -sombrero -soup bowl -space bar -space heater -space shuttle -spatula -speedboat -spider web, spider's web -spindle -sports car, sport car -spotlight, spot -stage -steam locomotive -steel arch bridge -steel drum -stethoscope -stole -stone wall -stopwatch, stop watch -stove -strainer -streetcar, tram, tramcar, trolley, trolley car -stretcher -studio couch, day bed -stupa, tope -submarine, pigboat, sub, U-boat -suit, suit of clothes -sundial -sunglass -sunglasses, dark glasses, shades -sunscreen, sunblock, sun blocker -suspension bridge -swab, swob, mop -sweatshirt -swimming trunks, bathing trunks -swing -switch, electric switch, electrical switch -syringe -table lamp -tank, army tank, armored combat vehicle, armoured combat vehicle -tape player -teapot -teddy, teddy bear -television, television system -tennis ball -thatch, thatched roof -theater curtain, theatre curtain -thimble -thresher, thrasher, threshing machine -throne -tile roof -toaster -tobacco shop, tobacconist shop, tobacconist -toilet seat -torch -totem pole -tow truck, tow car, wrecker -toyshop -tractor -trailer truck, tractor trailer, trucking rig, rig, articulated lorry, semi -tray -trench coat -tricycle, trike, velocipede -trimaran -tripod -triumphal arch -trolleybus, trolley coach, trackless trolley -trombone -tub, vat -turnstile -typewriter keyboard -umbrella -unicycle, monocycle -upright, upright piano -vacuum, vacuum cleaner -vase -vault -velvet -vending machine -vestment -viaduct -violin, fiddle -volleyball -waffle iron -wall clock -wallet, billfold, notecase, pocketbook -wardrobe, closet, press -warplane, military plane -washbasin, handbasin, washbowl, lavabo, wash-hand basin -washer, automatic washer, washing machine -water bottle -water jug -water tower -whiskey jug -whistle -wig -window screen -window shade -Windsor tie -wine bottle -wing -wok -wooden spoon -wool, woolen, woollen -worm fence, snake fence, snake-rail fence, Virginia fence -wreck -yawl -yurt -web site, website, internet site, site -comic book -crossword puzzle, crossword -street sign -traffic light, traffic signal, stoplight -book jacket, dust cover, dust jacket, dust wrapper -menu -plate -guacamole -consomme -hot pot, hotpot -trifle -ice cream, icecream -ice lolly, lolly, lollipop, popsicle -French loaf -bagel, beigel -pretzel -cheeseburger -hotdog, hot dog, red hot -mashed potato -head cabbage -broccoli -cauliflower -zucchini, courgette -spaghetti squash -acorn squash -butternut squash -cucumber, cuke -artichoke, globe artichoke -bell pepper -cardoon -mushroom -Granny Smith -strawberry -orange -lemon -fig -pineapple, ananas -banana -jackfruit, jak, jack -custard apple -pomegranate -hay -carbonara -chocolate sauce, chocolate syrup -dough -meat loaf, meatloaf -pizza, pizza pie -potpie -burrito -red wine -espresso -cup -eggnog -alp -bubble -cliff, drop, drop-off -coral reef -geyser -lakeside, lakeshore -promontory, headland, head, foreland -sandbar, sand bar -seashore, coast, seacoast, sea-coast -valley, vale -volcano -ballplayer, baseball player -groom, bridegroom -scuba diver -rapeseed -daisy -yellow lady's slipper, yellow lady-slipper, Cypripedium calceolus, Cypripedium parviflorum -corn -acorn -hip, rose hip, rosehip -buckeye, horse chestnut, conker -coral fungus -agaric -gyromitra -stinkhorn, carrion fungus -earthstar -hen-of-the-woods, hen of the woods, Polyporus frondosus, Grifola frondosa -bolete -ear, spike, capitulum -toilet tissue, toilet paper, bathroom tissue diff --git a/hub_module/modules/image/classification/efficientnetb5_imagenet/layers.py b/hub_module/modules/image/classification/efficientnetb5_imagenet/layers.py deleted file mode 100644 index 00f792b5..00000000 --- a/hub_module/modules/image/classification/efficientnetb5_imagenet/layers.py +++ /dev/null @@ -1,248 +0,0 @@ -# -*- coding:utf-8 -*- -# copyright (c) 2020 PaddlePaddle Authors. All Rights Reserve. -# -# Licensed under the Apache License, Version 2.0 (the "License"); -# you may not use this file except in compliance with the License. -# You may obtain a copy of the License at -# -# http://www.apache.org/licenses/LICENSE-2.0 -# -# Unless required by applicable law or agreed to in writing, software -# distributed under the License is distributed on an "AS IS" BASIS, -# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. -# See the License for the specific language governing permissions and -# limitations under the License. - -from __future__ import absolute_import -from __future__ import division -from __future__ import print_function - -import math -import warnings - -import paddle.fluid as fluid - - -def initial_type(name, - input, - op_type, - fan_out, - init="google", - use_bias=False, - filter_size=0, - stddev=0.02): - if init == "kaiming": - if op_type == 'conv': - fan_in = input.shape[1] * filter_size * filter_size - elif op_type == 'deconv': - fan_in = fan_out * filter_size * filter_size - else: - if len(input.shape) > 2: - fan_in = input.shape[1] * input.shape[2] * input.shape[3] - else: - fan_in = input.shape[1] - bound = 1 / math.sqrt(fan_in) - param_attr = fluid.ParamAttr( - name=name + "_weights", - initializer=fluid.initializer.Uniform(low=-bound, high=bound)) - if use_bias == True: - bias_attr = fluid.ParamAttr( - name=name + '_offset', - initializer=fluid.initializer.Uniform(low=-bound, high=bound)) - else: - bias_attr = False - elif init == 'google': - n = filter_size * filter_size * fan_out - param_attr = fluid.ParamAttr( - name=name + "_weights", - initializer=fluid.initializer.NormalInitializer( - loc=0.0, scale=math.sqrt(2.0 / n))) - if use_bias == True: - bias_attr = fluid.ParamAttr( - name=name + "_offset", - initializer=fluid.initializer.Constant(0.0)) - else: - bias_attr = False - - else: - param_attr = fluid.ParamAttr( - name=name + "_weights", - initializer=fluid.initializer.NormalInitializer( - loc=0.0, scale=stddev)) - if use_bias == True: - bias_attr = fluid.ParamAttr( - name=name + "_offset", - initializer=fluid.initializer.Constant(0.0)) - else: - bias_attr = False - return param_attr, bias_attr - - -def cal_padding(img_size, stride, filter_size, dilation=1): - """Calculate padding size.""" - if img_size % stride == 0: - out_size = max(filter_size - stride, 0) - else: - out_size = max(filter_size - (img_size % stride), 0) - return out_size // 2, out_size - out_size // 2 - - -def init_batch_norm_layer(name="batch_norm"): - param_attr = fluid.ParamAttr( - name=name + '_scale', initializer=fluid.initializer.Constant(1.0)) - bias_attr = fluid.ParamAttr( - name=name + '_offset', - initializer=fluid.initializer.Constant(value=0.0)) - return param_attr, bias_attr - - -def init_fc_layer(fout, name='fc'): - n = fout # fan-out - init_range = 1.0 / math.sqrt(n) - - param_attr = fluid.ParamAttr( - name=name + '_weights', - initializer=fluid.initializer.UniformInitializer( - low=-init_range, high=init_range)) - bias_attr = fluid.ParamAttr( - name=name + '_offset', - initializer=fluid.initializer.Constant(value=0.0)) - return param_attr, bias_attr - - -def norm_layer(input, norm_type='batch_norm', name=None): - if norm_type == 'batch_norm': - param_attr = fluid.ParamAttr( - name=name + '_weights', initializer=fluid.initializer.Constant(1.0)) - bias_attr = fluid.ParamAttr( - name=name + '_offset', - initializer=fluid.initializer.Constant(value=0.0)) - return fluid.layers.batch_norm( - input, - param_attr=param_attr, - bias_attr=bias_attr, - moving_mean_name=name + '_mean', - moving_variance_name=name + '_variance') - - elif norm_type == 'instance_norm': - helper = fluid.layer_helper.LayerHelper("instance_norm", **locals()) - dtype = helper.input_dtype() - epsilon = 1e-5 - mean = fluid.layers.reduce_mean(input, dim=[2, 3], keep_dim=True) - var = fluid.layers.reduce_mean( - fluid.layers.square(input - mean), dim=[2, 3], keep_dim=True) - if name is not None: - scale_name = name + "_scale" - offset_name = name + "_offset" - scale_param = fluid.ParamAttr( - name=scale_name, - initializer=fluid.initializer.Constant(1.0), - trainable=True) - offset_param = fluid.ParamAttr( - name=offset_name, - initializer=fluid.initializer.Constant(0.0), - trainable=True) - scale = helper.create_parameter( - attr=scale_param, shape=input.shape[1:2], dtype=dtype) - offset = helper.create_parameter( - attr=offset_param, shape=input.shape[1:2], dtype=dtype) - - tmp = fluid.layers.elementwise_mul(x=(input - mean), y=scale, axis=1) - tmp = tmp / fluid.layers.sqrt(var + epsilon) - tmp = fluid.layers.elementwise_add(tmp, offset, axis=1) - return tmp - else: - raise NotImplementedError("norm tyoe: [%s] is not support" % norm_type) - - -def conv2d(input, - num_filters=64, - filter_size=7, - stride=1, - stddev=0.02, - padding=0, - groups=None, - name="conv2d", - norm=None, - act=None, - relufactor=0.0, - use_bias=False, - padding_type=None, - initial="normal", - use_cudnn=True): - - if padding != 0 and padding_type != None: - warnings.warn( - 'padding value and padding type are set in the same time, and the final padding width and padding height are computed by padding_type' - ) - - param_attr, bias_attr = initial_type( - name=name, - input=input, - op_type='conv', - fan_out=num_filters, - init=initial, - use_bias=use_bias, - filter_size=filter_size, - stddev=stddev) - - def get_padding(filter_size, stride=1, dilation=1): - padding = ((stride - 1) + dilation * (filter_size - 1)) // 2 - return padding - - need_crop = False - if padding_type == "SAME": - top_padding, bottom_padding = cal_padding(input.shape[2], stride, - filter_size) - left_padding, right_padding = cal_padding(input.shape[2], stride, - filter_size) - height_padding = bottom_padding - width_padding = right_padding - if top_padding != bottom_padding or left_padding != right_padding: - height_padding = top_padding + stride - width_padding = left_padding + stride - need_crop = True - padding = [height_padding, width_padding] - elif padding_type == "VALID": - height_padding = 0 - width_padding = 0 - padding = [height_padding, width_padding] - elif padding_type == "DYNAMIC": - padding = get_padding(filter_size, stride) - else: - padding = padding - - conv = fluid.layers.conv2d( - input, - num_filters, - filter_size, - groups=groups, - name=name, - stride=stride, - padding=padding, - use_cudnn=use_cudnn, - param_attr=param_attr, - bias_attr=bias_attr) - - if need_crop: - conv = conv[:, :, 1:, 1:] - - if norm is not None: - conv = norm_layer(input=conv, norm_type=norm, name=name + "_norm") - if act == 'relu': - conv = fluid.layers.relu(conv, name=name + '_relu') - elif act == 'leaky_relu': - conv = fluid.layers.leaky_relu( - conv, alpha=relufactor, name=name + '_leaky_relu') - elif act == 'tanh': - conv = fluid.layers.tanh(conv, name=name + '_tanh') - elif act == 'sigmoid': - conv = fluid.layers.sigmoid(conv, name=name + '_sigmoid') - elif act == 'swish': - conv = fluid.layers.swish(conv, name=name + '_swish') - elif act == None: - conv = conv - else: - raise NotImplementedError("activation: [%s] is not support" % act) - - return conv diff --git a/hub_module/modules/image/classification/efficientnetb5_imagenet/module.py b/hub_module/modules/image/classification/efficientnetb5_imagenet/module.py index 9ee36ebd..bc6254be 100644 --- a/hub_module/modules/image/classification/efficientnetb5_imagenet/module.py +++ b/hub_module/modules/image/classification/efficientnetb5_imagenet/module.py @@ -1,4 +1,3 @@ -# -*- coding:utf-8 -*- # copyright (c) 2020 PaddlePaddle Authors. All Rights Reserve. # # Licensed under the Apache License, Version 2.0 (the "License"); @@ -13,312 +12,747 @@ # See the License for the specific language governing permissions and # limitations under the License. -from __future__ import absolute_import -from __future__ import division - -import ast -import argparse import os - -import numpy as np -import paddle.fluid as fluid -import paddlehub as hub -from paddle.fluid.core import PaddleTensor, AnalysisConfig, create_paddle_predictor -from paddlehub.module.module import moduleinfo, runnable, serving -from paddlehub.common.paddle_helper import add_vars_prefix - -from efficientnetb5_imagenet.processor import postprocess, base64_to_cv2 -from efficientnetb5_imagenet.data_feed import reader -from efficientnetb5_imagenet.efficientnet import EfficientNetB5 - - -@moduleinfo( - name="efficientnetb5_imagenet", - type="CV/image_classification", - author="paddlepaddle", - author_email="paddle-dev@baidu.com", - summary= - "EfficientNetB5 is a image classfication model, this module is trained with imagenet datasets.", - version="1.1.0") -class EfficientNetB5ImageNet(hub.Module): - def _initialize(self): - self.default_pretrained_model_path = os.path.join( - self.directory, "efficientnetb5_imagenet_infer_model") - label_file = os.path.join(self.directory, "label_list.txt") - with open(label_file, 'r', encoding='utf-8') as file: - self.label_list = file.read().split("\n")[:-1] - self.classification = self.classify - self._set_config() - - def get_expected_image_width(self): - return 224 - - def get_expected_image_height(self): - return 224 - - def get_pretrained_images_mean(self): - im_mean = np.array([0.485, 0.456, 0.406]).reshape(1, 3) - return im_mean - - def get_pretrained_images_std(self): - im_std = np.array([0.229, 0.224, 0.225]).reshape(1, 3) - return im_std - - def _set_config(self): - """ - predictor config setting +import math +import collections +import re +import copy + +import paddle +from paddle import ParamAttr +import paddle.nn as nn +import paddle.nn.functional as F +from paddle.nn import Conv2d, BatchNorm, Linear, Dropout +from paddle.nn import AdaptiveAvgPool2d, MaxPool2d, AvgPool2d +from paddlehub.module.module import moduleinfo +from paddlehub.module.cv_module import ImageClassifierModule + +GlobalParams = collections.namedtuple('GlobalParams', [ + 'batch_norm_momentum', + 'batch_norm_epsilon', + 'dropout_rate', + 'num_classes', + 'width_coefficient', + 'depth_coefficient', + 'depth_divisor', + 'min_depth', + 'drop_connect_rate', +]) + +BlockArgs = collections.namedtuple( + 'BlockArgs', + ['kernel_size', 'num_repeat', 'input_filters', 'output_filters', 'expand_ratio', 'id_skip', 'stride', 'se_ratio']) + +GlobalParams.__new__.__defaults__ = (None, ) * len(GlobalParams._fields) +BlockArgs.__new__.__defaults__ = (None, ) * len(BlockArgs._fields) + + +def efficientnet_params(model_name: str): + """ Map EfficientNet model name to parameter coefficients. """ + params_dict = { + # Coefficients: width,depth,resolution,dropout + 'efficientnet-b5': (1.6, 2.2, 456, 0.4) + } + return params_dict[model_name] + + +def efficientnet(width_coefficient: float = None, + depth_coefficient: float = None, + dropout_rate: float = 0.2, + drop_connect_rate: float = 0.2): + """ Get block arguments according to parameter and coefficients. """ + blocks_args = [ + 'r1_k3_s11_e1_i32_o16_se0.25', + 'r2_k3_s22_e6_i16_o24_se0.25', + 'r2_k5_s22_e6_i24_o40_se0.25', + 'r3_k3_s22_e6_i40_o80_se0.25', + 'r3_k5_s11_e6_i80_o112_se0.25', + 'r4_k5_s22_e6_i112_o192_se0.25', + 'r1_k3_s11_e6_i192_o320_se0.25', + ] + blocks_args = BlockDecoder.decode(blocks_args) + + global_params = GlobalParams(batch_norm_momentum=0.99, + batch_norm_epsilon=1e-3, + dropout_rate=dropout_rate, + drop_connect_rate=drop_connect_rate, + num_classes=1000, + width_coefficient=width_coefficient, + depth_coefficient=depth_coefficient, + depth_divisor=8, + min_depth=None) + + return blocks_args, global_params + + +def get_model_params(model_name: str, override_params: dict): + """ Get the block args and global params for a given model """ + if model_name.startswith('efficientnet'): + w, d, _, p = efficientnet_params(model_name) + blocks_args, global_params = efficientnet(width_coefficient=w, depth_coefficient=d, dropout_rate=p) + else: + raise NotImplementedError('model name is not pre-defined: %s' % model_name) + if override_params: + global_params = global_params._replace(**override_params) + return blocks_args, global_params + + +def round_filters(filters: int, global_params: dict): + """ Calculate and round number of filters based on depth multiplier. """ + multiplier = global_params.width_coefficient + if not multiplier: + return filters + divisor = global_params.depth_divisor + min_depth = global_params.min_depth + filters *= multiplier + min_depth = min_depth or divisor + new_filters = max(min_depth, int(filters + divisor / 2) // divisor * divisor) + if new_filters < 0.9 * filters: # prevent rounding by more than 10% + new_filters += divisor + return int(new_filters) + + +def round_repeats(repeats: int, global_params: dict): + """ Round number of filters based on depth multiplier. """ + multiplier = global_params.depth_coefficient + if not multiplier: + return repeats + return int(math.ceil(multiplier * repeats)) + + +class BlockDecoder(object): + """ + Block Decoder, straight from the official TensorFlow repository. + """ + @staticmethod + def _decode_block_string(block_string: str): + """ Gets a block through a string notation of arguments. """ + assert isinstance(block_string, str) + + ops = block_string.split('_') + options = {} + for op in ops: + splits = re.split(r'(\d.*)', op) + if len(splits) >= 2: + key, value = splits[:2] + options[key] = value + + # Check stride + cond_1 = ('s' in options and len(options['s']) == 1) + cond_2 = ((len(options['s']) == 2) and (options['s'][0] == options['s'][1])) + assert (cond_1 or cond_2) + + return BlockArgs(kernel_size=int(options['k']), + num_repeat=int(options['r']), + input_filters=int(options['i']), + output_filters=int(options['o']), + expand_ratio=int(options['e']), + id_skip=('noskip' not in block_string), + se_ratio=float(options['se']) if 'se' in options else None, + stride=[int(options['s'][0])]) + + @staticmethod + def _encode_block_string(block): + """Encodes a block to a string.""" + args = [ + 'r%d' % block.num_repeat, + 'k%d' % block.kernel_size, + 's%d%d' % (block.strides[0], block.strides[1]), + 'e%s' % block.expand_ratio, + 'i%d' % block.input_filters, + 'o%d' % block.output_filters + ] + if 0 < block.se_ratio <= 1: + args.append('se%s' % block.se_ratio) + if block.id_skip is False: + args.append('noskip') + return '_'.join(args) + + @staticmethod + def decode(string_list: list): """ - cpu_config = AnalysisConfig(self.default_pretrained_model_path) - cpu_config.disable_glog_info() - cpu_config.disable_gpu() - self.cpu_predictor = create_paddle_predictor(cpu_config) - - try: - _places = os.environ["CUDA_VISIBLE_DEVICES"] - int(_places[0]) - use_gpu = False - except: - use_gpu = True - if use_gpu: - gpu_config = AnalysisConfig(self.default_pretrained_model_path) - gpu_config.disable_glog_info() - gpu_config.enable_use_gpu( - memory_pool_init_size_mb=1000, device_id=0) - self.gpu_predictor = create_paddle_predictor(gpu_config) - - def context(self, - trainable=True, - pretrained=True, - override_params=None, - phase='train'): - """context for transfer learning. - - Args: - trainable (bool): Set parameters in program to be trainable. - pretrained (bool) : Whether to load pretrained model. - - Returns: - inputs (dict): key is 'image', corresponding vaule is image tensor. - outputs (dict): key is : - 'classification', corresponding value is the result of classification. - 'feature_map', corresponding value is the result of the layer before the fully connected layer. - context_prog (fluid.Program): program for transfer learning. - """ - if phase in ["dev", "test", "predict", "eval"]: - is_test = False - elif phase in ["train"]: - is_test = True - else: - raise ValueError( - "Phase %s is error, which must be one of train, dev, test, eval and predict." - % phase) - - context_prog = fluid.Program() - startup_prog = fluid.Program() - with fluid.program_guard(context_prog, startup_prog): - with fluid.unique_name.guard(): - image = fluid.layers.data( - name="image", shape=[3, 224, 224], dtype="float32") - efficientnet_b5 = EfficientNetB5( - override_params=override_params) - output, feature_map = efficientnet_b5.net( - input=image, - class_dim=len(self.label_list), - is_test=is_test) - - name_prefix = '@HUB_{}@'.format(self.name) - inputs = {'image': name_prefix + image.name} - outputs = { - 'classification': name_prefix + output.name, - 'feature_map': name_prefix + feature_map.name - } - add_vars_prefix(context_prog, name_prefix) - add_vars_prefix(startup_prog, name_prefix) - global_vars = context_prog.global_block().vars - inputs = { - key: global_vars[value] - for key, value in inputs.items() - } - outputs = { - key: global_vars[value] - for key, value in outputs.items() - } - - place = fluid.CPUPlace() - exe = fluid.Executor(place) - # pretrained - if pretrained: - - def _if_exist(var): - b = os.path.exists( - os.path.join(self.default_pretrained_model_path, - var.name)) - return b - - fluid.io.load_vars( - exe, - self.default_pretrained_model_path, - context_prog, - predicate=_if_exist) - else: - exe.run(startup_prog) - # trainable - for param in context_prog.global_block().iter_parameters(): - param.trainable = trainable - return inputs, outputs, context_prog - - def classify(self, - images=None, - paths=None, - batch_size=1, - use_gpu=False, - top_k=1): - """ - API for image classification. - - Args: - images (list[numpy.ndarray]): data of images, shape of each is [H, W, C], color space must be BGR. - paths (list[str]): The paths of images. - batch_size (int): batch size. - use_gpu (bool): Whether to use gpu. - top_k (int): Return top k results. + Decode a list of string notations to specify blocks in the network. - Returns: - res (list[dict]): The classfication results. + string_list: list of strings, each string is a notation of block + return + list of BlockArgs namedtuples of block args """ - if use_gpu: - try: - _places = os.environ["CUDA_VISIBLE_DEVICES"] - int(_places[0]) - except: - raise RuntimeError( - "Environment Variable CUDA_VISIBLE_DEVICES is not set correctly. If you wanna use gpu, please set CUDA_VISIBLE_DEVICES as cuda_device_id." - ) - - all_data = list() - for yield_data in reader(images, paths): - all_data.append(yield_data) - - total_num = len(all_data) - loop_num = int(np.ceil(total_num / batch_size)) - - res = list() - for iter_id in range(loop_num): - batch_data = list() - handle_id = iter_id * batch_size - for image_id in range(batch_size): - try: - batch_data.append(all_data[handle_id + image_id]) - except: - pass - # feed batch image - batch_image = np.array([data['image'] for data in batch_data]) - batch_image = PaddleTensor(batch_image.copy()) - predictor_output = self.gpu_predictor.run([ - batch_image - ]) if use_gpu else self.cpu_predictor.run([batch_image]) - out = postprocess( - data_out=predictor_output[0].as_ndarray(), - label_list=self.label_list, - top_k=top_k) - res += out - return res - - def save_inference_model(self, - dirname, - model_filename=None, - params_filename=None, - combined=True): - if combined: - model_filename = "__model__" if not model_filename else model_filename - params_filename = "__params__" if not params_filename else params_filename - place = fluid.CPUPlace() - exe = fluid.Executor(place) - - program, feeded_var_names, target_vars = fluid.io.load_inference_model( - dirname=self.default_pretrained_model_path, executor=exe) - - fluid.io.save_inference_model( - dirname=dirname, - main_program=program, - executor=exe, - feeded_var_names=feeded_var_names, - target_vars=target_vars, - model_filename=model_filename, - params_filename=params_filename) - - @serving - def serving_method(self, images, **kwargs): + assert isinstance(string_list, list) + blocks_args = [] + for block_string in string_list: + blocks_args.append(BlockDecoder._decode_block_string(block_string)) + return blocks_args + + @staticmethod + def encode(blocks_args: list): """ - Run as a service. - """ - images_decode = [base64_to_cv2(image) for image in images] - results = self.classify(images=images_decode, **kwargs) - return results + Encodes a list of BlockArgs to a list of strings. - @runnable - def run_cmd(self, argvs): - """ - Run as a command. + :param blocks_args: a list of BlockArgs namedtuples of block args + :return: a list of strings, each string is a notation of block """ - self.parser = argparse.ArgumentParser( - description="Run the {} module.".format(self.name), - prog='hub run {}'.format(self.name), - usage='%(prog)s', - add_help=True) - self.arg_input_group = self.parser.add_argument_group( - title="Input options", description="Input data. Required") - self.arg_config_group = self.parser.add_argument_group( - title="Config options", - description= - "Run configuration for controlling module behavior, not required.") - self.add_module_config_arg() - self.add_module_input_arg() - args = self.parser.parse_args(argvs) - results = self.classify( - paths=[args.input_path], - batch_size=args.batch_size, - use_gpu=args.use_gpu) - return results - - def add_module_config_arg(self): - """ - Add the command config options. - """ - self.arg_config_group.add_argument( - '--use_gpu', - type=ast.literal_eval, - default=False, - help="whether use GPU or not.") - self.arg_config_group.add_argument( - '--batch_size', - type=ast.literal_eval, - default=1, - help="batch size.") - self.arg_config_group.add_argument( - '--top_k', - type=ast.literal_eval, - default=1, - help="Return top k results.") - - def add_module_input_arg(self): - """ - Add the command input options. - """ - self.arg_input_group.add_argument( - '--input_path', type=str, help="path to image.") - - -if __name__ == '__main__': - b5 = EfficientNetB5ImageNet() - b5.context() - import cv2 - test_image = [cv2.imread('dog.jpeg')] - res = b5.classification(images=test_image) - print(res) - res = b5.classification(paths=['dog.jpeg']) - print(res) - res = b5.classification(images=test_image) - print(res) - res = b5.classify(images=test_image) - print(res) + block_strings = [] + for block in blocks_args: + block_strings.append(BlockDecoder._encode_block_string(block)) + return block_strings + + +def initial_type(name: str, use_bias: bool = False): + param_attr = ParamAttr(name=name + "_weights") + if use_bias: + bias_attr = ParamAttr(name=name + "_offset") + else: + bias_attr = False + return param_attr, bias_attr + + +def init_batch_norm_layer(name: str = "batch_norm"): + param_attr = ParamAttr(name=name + "_scale") + bias_attr = ParamAttr(name=name + "_offset") + return param_attr, bias_attr + + +def init_fc_layer(name: str = "fc"): + param_attr = ParamAttr(name=name + "_weights") + bias_attr = ParamAttr(name=name + "_offset") + return param_attr, bias_attr + + +def cal_padding(img_size: int, stride: int, filter_size: int, dilation: int = 1): + """Calculate padding size.""" + if img_size % stride == 0: + out_size = max(filter_size - stride, 0) + else: + out_size = max(filter_size - (img_size % stride), 0) + return out_size // 2, out_size - out_size // 2 + + +inp_shape = {"b5": [456, 228, 228, 114, 57, 29, 29, 15]} + + +def _drop_connect(inputs: paddle.Tensor, prob: float, is_test: bool): + """Drop input connection""" + if is_test: + return inputs + keep_prob = 1.0 - prob + inputs_shape = paddle.shape(inputs) + random_tensor = keep_prob + paddle.rand(shape=[inputs_shape[0], 1, 1, 1]) + binary_tensor = paddle.floor(random_tensor) + output = inputs / keep_prob * binary_tensor + return output + + +class Conv2ds(nn.Layer): + """Basic conv layer""" + def __init__(self, + input_channels: int, + output_channels: int, + filter_size: int, + stride: int = 1, + padding: int = 0, + groups: int = None, + name: str = "conv2d", + act: str = None, + use_bias: bool = False, + padding_type: str = None, + model_name: str = None, + cur_stage: str = None): + super(Conv2ds, self).__init__() + assert act in [None, "swish", "sigmoid"] + self.act = act + + param_attr, bias_attr = initial_type(name=name, use_bias=use_bias) + + def get_padding(filter_size, stride=1, dilation=1): + padding = ((stride - 1) + dilation * (filter_size - 1)) // 2 + return padding + + inps = 1 if model_name == None and cur_stage == None else inp_shape[model_name][cur_stage] + self.need_crop = False + if padding_type == "SAME": + top_padding, bottom_padding = cal_padding(inps, stride, filter_size) + left_padding, right_padding = cal_padding(inps, stride, filter_size) + height_padding = bottom_padding + width_padding = right_padding + if top_padding != bottom_padding or left_padding != right_padding: + height_padding = top_padding + stride + width_padding = left_padding + stride + self.need_crop = True + padding = [height_padding, width_padding] + elif padding_type == "VALID": + height_padding = 0 + width_padding = 0 + padding = [height_padding, width_padding] + elif padding_type == "DYNAMIC": + padding = get_padding(filter_size, stride) + else: + padding = padding_type + + groups = 1 if groups is None else groups + self._conv = Conv2d(input_channels, + output_channels, + filter_size, + groups=groups, + stride=stride, + padding=padding, + weight_attr=param_attr, + bias_attr=bias_attr) + + def forward(self, inputs: paddle.Tensor): + x = self._conv(inputs) + if self.act == "swish": + x = F.swish(x) + elif self.act == "sigmoid": + x = F.sigmoid(x) + + if self.need_crop: + x = x[:, :, 1:, 1:] + return x + + +class ConvBNLayer(nn.Layer): + """Basic conv bn layer.""" + def __init__(self, + input_channels: int, + filter_size: int, + output_channels: int, + stride: int = 1, + num_groups: int = 1, + padding_type: str = "SAME", + conv_act: str = None, + bn_act: str = "swish", + use_bn: bool = True, + use_bias: bool = False, + name: str = None, + conv_name: str = None, + bn_name: str = None, + model_name: str = None, + cur_stage: str = None): + super(ConvBNLayer, self).__init__() + + self._conv = Conv2ds(input_channels=input_channels, + output_channels=output_channels, + filter_size=filter_size, + stride=stride, + groups=num_groups, + act=conv_act, + padding_type=padding_type, + name=conv_name, + use_bias=use_bias, + model_name=model_name, + cur_stage=cur_stage) + self.use_bn = use_bn + if use_bn is True: + bn_name = name + bn_name + param_attr, bias_attr = init_batch_norm_layer(bn_name) + + self._bn = BatchNorm(num_channels=output_channels, + act=bn_act, + momentum=0.99, + epsilon=0.001, + moving_mean_name=bn_name + "_mean", + moving_variance_name=bn_name + "_variance", + param_attr=param_attr, + bias_attr=bias_attr) + + def forward(self, inputs: paddle.Tensor): + if self.use_bn: + x = self._conv(inputs) + x = self._bn(x) + return x + else: + return self._conv(inputs) + + +class ExpandConvNorm(nn.Layer): + """Expand conv norm layer.""" + def __init__(self, + input_channels: int, + block_args: dict, + padding_type: str, + name: str = None, + model_name: str = None, + cur_stage: str = None): + super(ExpandConvNorm, self).__init__() + + self.oup = block_args.input_filters * block_args.expand_ratio + self.expand_ratio = block_args.expand_ratio + + if self.expand_ratio != 1: + self._conv = ConvBNLayer(input_channels, + 1, + self.oup, + bn_act=None, + padding_type=padding_type, + name=name, + conv_name=name + "_expand_conv", + bn_name="_bn0", + model_name=model_name, + cur_stage=cur_stage) + + def forward(self, inputs: paddle.Tensor): + if self.expand_ratio != 1: + return self._conv(inputs) + else: + return inputs + + +class DepthwiseConvNorm(nn.Layer): + """Depthwise conv norm layer.""" + def __init__(self, + input_channels: int, + block_args: dict, + padding_type: str, + name: str = None, + model_name: str = None, + cur_stage: str = None): + super(DepthwiseConvNorm, self).__init__() + + self.k = block_args.kernel_size + self.s = block_args.stride + if isinstance(self.s, list) or isinstance(self.s, tuple): + self.s = self.s[0] + oup = block_args.input_filters * block_args.expand_ratio + + self._conv = ConvBNLayer(input_channels, + self.k, + oup, + self.s, + num_groups=input_channels, + bn_act=None, + padding_type=padding_type, + name=name, + conv_name=name + "_depthwise_conv", + bn_name="_bn1", + model_name=model_name, + cur_stage=cur_stage) + + def forward(self, inputs: paddle.Tensor): + return self._conv(inputs) + + +class ProjectConvNorm(nn.Layer): + """Projection conv bn layer.""" + def __init__(self, + input_channels: int, + block_args: dict, + padding_type: str, + name: str = None, + model_name: str = None, + cur_stage: str = None): + super(ProjectConvNorm, self).__init__() + + final_oup = block_args.output_filters + + self._conv = ConvBNLayer(input_channels, + 1, + final_oup, + bn_act=None, + padding_type=padding_type, + name=name, + conv_name=name + "_project_conv", + bn_name="_bn2", + model_name=model_name, + cur_stage=cur_stage) + + def forward(self, inputs: paddle.Tensor): + return self._conv(inputs) + + +class SEBlock(nn.Layer): + """Basic Squeeze-and-Excitation block for Efficientnet.""" + def __init__(self, + input_channels: int, + num_squeezed_channels: int, + oup: int, + padding_type: str, + name: str = None, + model_name: str = None, + cur_stage: str = None): + super(SEBlock, self).__init__() + + self._pool = AdaptiveAvgPool2d(1) + self._conv1 = Conv2ds(input_channels, + num_squeezed_channels, + 1, + use_bias=True, + padding_type=padding_type, + act="swish", + name=name + "_se_reduce") + + self._conv2 = Conv2ds(num_squeezed_channels, + oup, + 1, + act="sigmoid", + use_bias=True, + padding_type=padding_type, + name=name + "_se_expand") + + def forward(self, inputs: paddle.Tensor): + x = self._pool(inputs) + x = self._conv1(x) + x = self._conv2(x) + return paddle.multiply(inputs, x) + + +class MbConvBlock(nn.Layer): + """Mobile inverted bottleneck convolution for Efficientnet.""" + def __init__(self, + input_channels: int, + block_args: dict, + padding_type: str, + use_se: bool, + name: str = None, + drop_connect_rate: float = None, + is_test: bool = False, + model_name: str = None, + cur_stage: str = None): + super(MbConvBlock, self).__init__() + + oup = block_args.input_filters * block_args.expand_ratio + self.block_args = block_args + self.has_se = use_se and (block_args.se_ratio is not None) and (0 < block_args.se_ratio <= 1) + self.id_skip = block_args.id_skip + self.expand_ratio = block_args.expand_ratio + self.drop_connect_rate = drop_connect_rate + self.is_test = is_test + + if self.expand_ratio != 1: + self._ecn = ExpandConvNorm(input_channels, + block_args, + padding_type=padding_type, + name=name, + model_name=model_name, + cur_stage=cur_stage) + + self._dcn = DepthwiseConvNorm(input_channels * block_args.expand_ratio, + block_args, + padding_type=padding_type, + name=name, + model_name=model_name, + cur_stage=cur_stage) + + if self.has_se: + num_squeezed_channels = max(1, int(block_args.input_filters * block_args.se_ratio)) + self._se = SEBlock(input_channels * block_args.expand_ratio, + num_squeezed_channels, + oup, + padding_type=padding_type, + name=name, + model_name=model_name, + cur_stage=cur_stage) + + self._pcn = ProjectConvNorm(input_channels * block_args.expand_ratio, + block_args, + padding_type=padding_type, + name=name, + model_name=model_name, + cur_stage=cur_stage) + + def forward(self, inputs: paddle.Tensor): + x = inputs + if self.expand_ratio != 1: + x = self._ecn(x) + x = F.swish(x) + x = self._dcn(x) + x = F.swish(x) + if self.has_se: + x = self._se(x) + x = self._pcn(x) + if self.id_skip and \ + self.block_args.stride == 1 and \ + self.block_args.input_filters == self.block_args.output_filters: + if self.drop_connect_rate: + x = _drop_connect(x, self.drop_connect_rate, self.is_test) + x = paddle.elementwise_add(x, inputs) + return x + + +class ConvStemNorm(nn.Layer): + """Basic conv stem norm block for extracting features.""" + def __init__(self, + input_channels: int, + padding_type: str, + _global_params: dict, + name: str = None, + model_name: str = None, + cur_stage: str = None): + super(ConvStemNorm, self).__init__() + + output_channels = round_filters(32, _global_params) + self._conv = ConvBNLayer(input_channels, + filter_size=3, + output_channels=output_channels, + stride=2, + bn_act=None, + padding_type=padding_type, + name="", + conv_name="_conv_stem", + bn_name="_bn0", + model_name=model_name, + cur_stage=cur_stage) + + def forward(self, inputs: paddle.Tensor): + return self._conv(inputs) + + +class ExtractFeatures(nn.Layer): + """Extract features.""" + def __init__(self, + input_channels: int, + _block_args: dict, + _global_params: dict, + padding_type: str, + use_se: bool, + is_test: bool, + model_name: str = None): + super(ExtractFeatures, self).__init__() + + self._global_params = _global_params + + self._conv_stem = ConvStemNorm(input_channels, + padding_type=padding_type, + _global_params=_global_params, + model_name=model_name, + cur_stage=0) + + self.block_args_copy = copy.deepcopy(_block_args) + idx = 0 + block_size = 0 + for block_arg in self.block_args_copy: + block_arg = block_arg._replace(input_filters=round_filters(block_arg.input_filters, _global_params), + output_filters=round_filters(block_arg.output_filters, _global_params), + num_repeat=round_repeats(block_arg.num_repeat, _global_params)) + block_size += 1 + for _ in range(block_arg.num_repeat - 1): + block_size += 1 + + self.conv_seq = [] + cur_stage = 1 + for block_args in _block_args: + block_args = block_args._replace(input_filters=round_filters(block_args.input_filters, _global_params), + output_filters=round_filters(block_args.output_filters, _global_params), + num_repeat=round_repeats(block_args.num_repeat, _global_params)) + + drop_connect_rate = self._global_params.drop_connect_rate if not is_test else 0 + if drop_connect_rate: + drop_connect_rate *= float(idx) / block_size + + _mc_block = self.add_sublayer( + "_blocks." + str(idx) + ".", + MbConvBlock(block_args.input_filters, + block_args=block_args, + padding_type=padding_type, + use_se=use_se, + name="_blocks." + str(idx) + ".", + drop_connect_rate=drop_connect_rate, + model_name=model_name, + cur_stage=cur_stage)) + self.conv_seq.append(_mc_block) + idx += 1 + if block_args.num_repeat > 1: + block_args = block_args._replace(input_filters=block_args.output_filters, stride=1) + for _ in range(block_args.num_repeat - 1): + drop_connect_rate = self._global_params.drop_connect_rate if not is_test else 0 + if drop_connect_rate: + drop_connect_rate *= float(idx) / block_size + _mc_block = self.add_sublayer( + "block." + str(idx) + ".", + MbConvBlock(block_args.input_filters, + block_args, + padding_type=padding_type, + use_se=use_se, + name="_blocks." + str(idx) + ".", + drop_connect_rate=drop_connect_rate, + model_name=model_name, + cur_stage=cur_stage)) + self.conv_seq.append(_mc_block) + idx += 1 + cur_stage += 1 + + def forward(self, inputs: paddle.Tensor): + x = self._conv_stem(inputs) + x = F.swish(x) + for _mc_block in self.conv_seq: + x = _mc_block(x) + return x + + +@moduleinfo(name="efficientnetb5_imagenet", + type="cv/classification", + author="paddlepaddle", + author_email="", + summary="efficientnetb5_imagenet is a classification model, " + "this module is trained with Imagenet dataset.", + version="1.1.0", + meta=ImageClassifierModule) +class EfficientNet_B5(nn.Layer): + def __init__(self, + is_test: bool = False, + padding_type: str = "SAME", + override_params: dict = None, + use_se: bool = True, + class_dim: int = 1000, + load_checkpoint: str = None): + super(EfficientNet_B5, self).__init__() + + model_name = 'efficientnet-b5' + self.name = "b5" + self._block_args, self._global_params = get_model_params(model_name, override_params) + self.padding_type = padding_type + self.use_se = use_se + self.is_test = is_test + + self._ef = ExtractFeatures(3, + self._block_args, + self._global_params, + self.padding_type, + self.use_se, + self.is_test, + model_name=self.name) + + output_channels = round_filters(1280, self._global_params) + oup = 512 + + self._conv = ConvBNLayer(oup, + 1, + output_channels, + bn_act="swish", + padding_type=self.padding_type, + name="", + conv_name="_conv_head", + bn_name="_bn1", + model_name=self.name, + cur_stage=7) + self._pool = AdaptiveAvgPool2d(1) + + if self._global_params.dropout_rate: + self._drop = Dropout(p=self._global_params.dropout_rate, mode="upscale_in_train") + + param_attr, bias_attr = init_fc_layer("_fc") + self._fc = Linear(output_channels, class_dim, weight_attr=param_attr, bias_attr=bias_attr) + + if load_checkpoint is not None: + model_dict = paddle.load(load_checkpoint)[0] + self.set_dict(model_dict) + print("load custom checkpoint success") + + else: + checkpoint = os.path.join(self.directory, 'efficientnet_b5_imagenet.pdparams') + if not os.path.exists(checkpoint): + os.system( + 'wget https://paddlehub.bj.bcebos.com/dygraph/image_classification/efficientnet_b5_imagenet.pdparams -O ' + + checkpoint) + model_dict = paddle.load(checkpoint)[0] + self.set_dict(model_dict) + print("load pretrained checkpoint success") + + def forward(self, inputs: paddle.Tensor): + x = self._ef(inputs) + x = self._conv(x) + x = self._pool(x) + if self._global_params.dropout_rate: + x = self._drop(x) + x = paddle.squeeze(x, axis=[2, 3]) + x = self._fc(x) + return x diff --git a/hub_module/modules/image/classification/efficientnetb5_imagenet/processor.py b/hub_module/modules/image/classification/efficientnetb5_imagenet/processor.py deleted file mode 100644 index 12ccc9b2..00000000 --- a/hub_module/modules/image/classification/efficientnetb5_imagenet/processor.py +++ /dev/null @@ -1,69 +0,0 @@ -# -*- coding:utf-8 -*- -# copyright (c) 2020 PaddlePaddle Authors. All Rights Reserve. -# -# Licensed under the Apache License, Version 2.0 (the "License"); -# you may not use this file except in compliance with the License. -# You may obtain a copy of the License at -# -# http://www.apache.org/licenses/LICENSE-2.0 -# -# Unless required by applicable law or agreed to in writing, software -# distributed under the License is distributed on an "AS IS" BASIS, -# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. -# See the License for the specific language governing permissions and -# limitations under the License. - -from __future__ import absolute_import -from __future__ import division -from __future__ import print_function - -import base64 -import cv2 -import os - -import numpy as np - - -def base64_to_cv2(b64str): - data = base64.b64decode(b64str.encode('utf8')) - data = np.fromstring(data, np.uint8) - data = cv2.imdecode(data, cv2.IMREAD_COLOR) - return data - - -def softmax(x): - orig_shape = x.shape - if len(x.shape) > 1: - tmp = np.max(x, axis=1) - x -= tmp.reshape((x.shape[0], 1)) - x = np.exp(x) - tmp = np.sum(x, axis=1) - x /= tmp.reshape((x.shape[0], 1)) - else: - tmp = np.max(x) - x -= tmp - x = np.exp(x) - tmp = np.sum(x) - x /= tmp - return x - - -def postprocess(data_out, label_list, top_k): - """ - Postprocess output of network, one image at a time. - - Args: - data_out (numpy.ndarray): output data of network. - label_list (list): list of label. - top_k (int): Return top k results. - """ - output = [] - for result in data_out: - result_i = softmax(result) - output_i = {} - indexs = np.argsort(result_i)[::-1][0:top_k] - for index in indexs: - label = label_list[index].split(',')[0] - output_i[label] = float(result_i[index]) - output.append(output_i) - return output diff --git a/hub_module/modules/image/classification/efficientnetb6_imagenet/README.md b/hub_module/modules/image/classification/efficientnetb6_imagenet/README.md deleted file mode 100644 index da687d10..00000000 --- a/hub_module/modules/image/classification/efficientnetb6_imagenet/README.md +++ /dev/null @@ -1,149 +0,0 @@ -## 命令行预测 - -``` -hub run efficientnetb6_imagenet --input_path "/PATH/TO/IMAGE" -``` - -## API - -```python -def get_expected_image_width() -``` - -返回预处理的图片宽度，也就是224。 - -```python -def get_expected_image_height() -``` - -返回预处理的图片高度，也就是224。 - -```python -def get_pretrained_images_mean() -``` - -返回预处理的图片均值，也就是 \[0.485, 0.456, 0.406\]。 - -```python -def get_pretrained_images_std() -``` - -返回预处理的图片标准差，也就是 \[0.229, 0.224, 0.225\]。 - - -```python -def context(trainable=True, pretrained=True) -``` - -**参数** - -* trainable (bool): 计算图的参数是否为可训练的； -* pretrained (bool): 是否加载默认的预训练模型。 - -**返回** - -* inputs (dict): 计算图的输入，key 为 'image', value 为图片的张量； -* outputs (dict): 计算图的输出，key 为 'classification' 和 'feature_map'，其相应的值为： - * classification (paddle.fluid.framework.Variable): 分类结果，也就是全连接层的输出； - * feature\_map (paddle.fluid.framework.Variable): 特征匹配，全连接层前面的那个张量。 -* context\_prog(fluid.Program): 计算图，用于迁移学习。 - -```python -def classify(images=None, - paths=None, - batch_size=1, - use_gpu=False, - top_k=1): -``` - -**参数** - -* images (list\[numpy.ndarray\]): 图片数据，每一个图片数据的shape 均为 \[H, W, C\]，颜色空间为 BGR； -* paths (list\[str\]): 图片的路径； -* batch\_size (int): batch 的大小； -* use\_gpu (bool): 是否使用 GPU 来预测； -* top\_k (int): 返回预测结果的前 k 个。 - -**返回** - -res (list\[dict\]): 分类结果，列表的每一个元素均为字典，其中 key 为识别动物的类别，value为置信度。 - -```python -def save_inference_model(dirname, - model_filename=None, - params_filename=None, - combined=True) -``` - -将模型保存到指定路径。 - -**参数** - -* dirname: 存在模型的目录名称 -* model\_filename: 模型文件名称，默认为\_\_model\_\_ -* params\_filename: 参数文件名称，默认为\_\_params\_\_(仅当`combined`为True时生效) -* combined: 是否将参数保存到统一的一个文件中 - -## 代码示例 - -```python -import paddlehub as hub -import cv2 - -classifier = hub.Module(name="efficientnetb6_imagenet") - -result = classifier.classify(images=[cv2.imread('/PATH/TO/IMAGE')]) -# or -# result = classifier.classify(paths=['/PATH/TO/IMAGE']) -``` - -## 服务部署 - -PaddleHub Serving可以部署一个在线图像识别服务。 - -## 第一步：启动PaddleHub Serving - -运行启动命令： -```shell -$ hub serving start -m efficientnetb6_imagenet -``` - -这样就完成了一个在线图像识别服务化API的部署，默认端口号为8866。 - -**NOTE:** 如使用GPU预测，则需要在启动服务之前，请设置CUDA\_VISIBLE\_DEVICES环境变量，否则不用设置。 - -## 第二步：发送预测请求 - -配置好服务端，以下数行代码即可实现发送预测请求，获取预测结果 - -```python -import requests -import json -import cv2 -import base64 - - -def cv2_to_base64(image): - data = cv2.imencode('.jpg', image)[1] - return base64.b64encode(data.tostring()).decode('utf8') - - -# 发送HTTP请求 -data = {'images':[cv2_to_base64(cv2.imread("/PATH/TO/IMAGE"))]} -headers = {"Content-type": "application/json"} -url = "http://127.0.0.1:8866/predict/efficientnetb6_imagenet" -r = requests.post(url=url, headers=headers, data=json.dumps(data)) - -# 打印预测结果 -print(r.json()["results"]) -``` - -### 查看代码 - -https://github.com/PaddlePaddle/PaddleClas - -### 依赖 - -paddlepaddle >= 1.6.2 - -paddlehub >= 1.6.0 diff --git a/hub_module/modules/image/classification/efficientnetb6_imagenet/__init__.py b/hub_module/modules/image/classification/efficientnetb6_imagenet/__init__.py deleted file mode 100644 index 19a990da..00000000 --- a/hub_module/modules/image/classification/efficientnetb6_imagenet/__init__.py +++ /dev/null @@ -1 +0,0 @@ -# -*- coding:utf-8 -*- diff --git a/hub_module/modules/image/classification/efficientnetb6_imagenet/data_feed.py b/hub_module/modules/image/classification/efficientnetb6_imagenet/data_feed.py deleted file mode 100644 index 11181458..00000000 --- a/hub_module/modules/image/classification/efficientnetb6_imagenet/data_feed.py +++ /dev/null @@ -1,98 +0,0 @@ -# -*- coding:utf-8 -*- -# copyright (c) 2020 PaddlePaddle Authors. All Rights Reserve. -# -# Licensed under the Apache License, Version 2.0 (the "License"); -# you may not use this file except in compliance with the License. -# You may obtain a copy of the License at -# -# http://www.apache.org/licenses/LICENSE-2.0 -# -# Unless required by applicable law or agreed to in writing, software -# distributed under the License is distributed on an "AS IS" BASIS, -# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. -# See the License for the specific language governing permissions and -# limitations under the License. - -import os -import time -from collections import OrderedDict - -import cv2 -import numpy as np -from PIL import Image - -__all__ = ['reader'] - -DATA_DIM = 224 -img_mean = np.array([0.485, 0.456, 0.406]).reshape((3, 1, 1)) -img_std = np.array([0.229, 0.224, 0.225]).reshape((3, 1, 1)) - - -def resize_short(img, target_size): - percent = float(target_size) / min(img.size[0], img.size[1]) - resized_width = int(round(img.size[0] * percent)) - resized_height = int(round(img.size[1] * percent)) - img = img.resize((resized_width, resized_height), Image.LANCZOS) - return img - - -def crop_image(img, target_size, center): - width, height = img.size - size = target_size - if center == True: - w_start = (width - size) / 2 - h_start = (height - size) / 2 - else: - w_start = np.random.randint(0, width - size + 1) - h_start = np.random.randint(0, height - size + 1) - w_end = w_start + size - h_end = h_start + size - img = img.crop((w_start, h_start, w_end, h_end)) - return img - - -def process_image(img): - img = resize_short(img, target_size=256) - img = crop_image(img, target_size=DATA_DIM, center=True) - if img.mode != 'RGB': - img = img.convert('RGB') - img = np.array(img).astype('float32').transpose((2, 0, 1)) / 255 - img -= img_mean - img /= img_std - return img - - -def reader(images=None, paths=None): - """ - Preprocess to yield image. - - Args: - images (list[numpy.ndarray]): images data, shape of each is [H, W, C]. - paths (list[str]): paths to images. - - Yield: - each (collections.OrderedDict): info of original image, preprocessed image. - """ - component = list() - if paths: - for im_path in paths: - each = OrderedDict() - assert os.path.isfile( - im_path), "The {} isn't a valid file path.".format(im_path) - each['org_im_path'] = im_path - each['org_im'] = Image.open(im_path) - each['org_im_width'], each['org_im_height'] = each['org_im'].size - component.append(each) - if images is not None: - assert type(images), "images is a list." - for im in images: - each = OrderedDict() - each['org_im'] = Image.fromarray(im[:, :, ::-1]) - each['org_im_path'] = 'ndarray_time={}'.format( - round(time.time(), 6) * 1e6) - each['org_im_width'], each['org_im_height'] = each['org_im'].size - component.append(each) - - for element in component: - element['image'] = process_image(element['org_im']) - yield element diff --git a/hub_module/modules/image/classification/efficientnetb6_imagenet/efficientnet.py b/hub_module/modules/image/classification/efficientnetb6_imagenet/efficientnet.py deleted file mode 100644 index e79f48c1..00000000 --- a/hub_module/modules/image/classification/efficientnetb6_imagenet/efficientnet.py +++ /dev/null @@ -1,623 +0,0 @@ -# -*- coding:utf-8 -*- -# copyright (c) 2020 PaddlePaddle Authors. All Rights Reserve. -# -# Licensed under the Apache License, Version 2.0 (the "License"); -# you may not use this file except in compliance with the License. -# You may obtain a copy of the License at -# -# http://www.apache.org/licenses/LICENSE-2.0 -# -# Unless required by applicable law or agreed to in writing, software -# distributed under the License is distributed on an "AS IS" BASIS, -# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. -# See the License for the specific language governing permissions and -# limitations under the License. - -from __future__ import absolute_import -from __future__ import division -from __future__ import print_function - -import collections -import re -import math -import copy - -import paddle.fluid as fluid -from efficientnetb6_imagenet.layers import conv2d, init_batch_norm_layer, init_fc_layer - -__all__ = [ - 'EfficientNet', 'EfficientNetB0_small', 'EfficientNetB0', 'EfficientNetB1', - 'EfficientNetB2', 'EfficientNetB3', 'EfficientNetB4', 'EfficientNetB5', - 'EfficientNetB6', 'EfficientNetB7' -] - -GlobalParams = collections.namedtuple('GlobalParams', [ - 'batch_norm_momentum', - 'batch_norm_epsilon', - 'dropout_rate', - 'num_classes', - 'width_coefficient', - 'depth_coefficient', - 'depth_divisor', - 'min_depth', - 'drop_connect_rate', -]) - -BlockArgs = collections.namedtuple('BlockArgs', [ - 'kernel_size', 'num_repeat', 'input_filters', 'output_filters', - 'expand_ratio', 'id_skip', 'stride', 'se_ratio' -]) - -GlobalParams.__new__.__defaults__ = (None, ) * len(GlobalParams._fields) -BlockArgs.__new__.__defaults__ = (None, ) * len(BlockArgs._fields) - - -def efficientnet_params(model_name): - """ Map EfficientNet model name to parameter coefficients. """ - params_dict = { - # Coefficients: width,depth,resolution,dropout - 'efficientnet-b0': (1.0, 1.0, 224, 0.2), - 'efficientnet-b1': (1.0, 1.1, 240, 0.2), - 'efficientnet-b2': (1.1, 1.2, 260, 0.3), - 'efficientnet-b3': (1.2, 1.4, 300, 0.3), - 'efficientnet-b4': (1.4, 1.8, 380, 0.4), - 'efficientnet-b5': (1.6, 2.2, 456, 0.4), - 'efficientnet-b6': (1.8, 2.6, 528, 0.5), - 'efficientnet-b7': (2.0, 3.1, 600, 0.5), - } - return params_dict[model_name] - - -def efficientnet(width_coefficient=None, - depth_coefficient=None, - dropout_rate=0.2, - drop_connect_rate=0.2): - """ Get block arguments according to parameter and coefficients. """ - blocks_args = [ - 'r1_k3_s11_e1_i32_o16_se0.25', - 'r2_k3_s22_e6_i16_o24_se0.25', - 'r2_k5_s22_e6_i24_o40_se0.25', - 'r3_k3_s22_e6_i40_o80_se0.25', - 'r3_k5_s11_e6_i80_o112_se0.25', - 'r4_k5_s22_e6_i112_o192_se0.25', - 'r1_k3_s11_e6_i192_o320_se0.25', - ] - blocks_args = BlockDecoder.decode(blocks_args) - - global_params = GlobalParams( - batch_norm_momentum=0.99, - batch_norm_epsilon=1e-3, - dropout_rate=dropout_rate, - drop_connect_rate=drop_connect_rate, - num_classes=1000, - width_coefficient=width_coefficient, - depth_coefficient=depth_coefficient, - depth_divisor=8, - min_depth=None) - - return blocks_args, global_params - - -def get_model_params(model_name, override_params): - """ Get the block args and global params for a given model """ - if model_name.startswith('efficientnet'): - w, d, _, p = efficientnet_params(model_name) - blocks_args, global_params = efficientnet( - width_coefficient=w, depth_coefficient=d, dropout_rate=p) - else: - raise NotImplementedError( - 'model name is not pre-defined: %s' % model_name) - if override_params: - global_params = global_params._replace(**override_params) - return blocks_args, global_params - - -def round_filters(filters, global_params): - """ Calculate and round number of filters based on depth multiplier. """ - multiplier = global_params.width_coefficient - if not multiplier: - return filters - divisor = global_params.depth_divisor - min_depth = global_params.min_depth - filters *= multiplier - min_depth = min_depth or divisor - new_filters = max(min_depth, - int(filters + divisor / 2) // divisor * divisor) - if new_filters < 0.9 * filters: # prevent rounding by more than 10% - new_filters += divisor - return int(new_filters) - - -def round_repeats(repeats, global_params): - """ Round number of filters based on depth multiplier. """ - multiplier = global_params.depth_coefficient - if not multiplier: - return repeats - return int(math.ceil(multiplier * repeats)) - - -class EfficientNet(): - def __init__(self, - name='b0', - padding_type='SAME', - override_params=None, - is_test=False, - use_se=True): - valid_names = ['b' + str(i) for i in range(8)] - assert name in valid_names, 'efficient name should be in b0~b7' - model_name = 'efficientnet-' + name - self._blocks_args, self._global_params = get_model_params( - model_name, override_params) - self._bn_mom = self._global_params.batch_norm_momentum - self._bn_eps = self._global_params.batch_norm_epsilon - self.padding_type = padding_type - self.use_se = use_se - - def net(self, input, class_dim=1000, is_test=False): - - conv = self.extract_features(input, is_test=is_test) - - out_channels = round_filters(1280, self._global_params) - conv = self.conv_bn_layer( - conv, - num_filters=out_channels, - filter_size=1, - bn_act='swish', - bn_mom=self._bn_mom, - bn_eps=self._bn_eps, - padding_type=self.padding_type, - name='', - conv_name='_conv_head', - bn_name='_bn1') - - pool = fluid.layers.pool2d( - input=conv, pool_type='avg', global_pooling=True, use_cudnn=False) - - if not is_test and self._global_params.dropout_rate: - pool = fluid.layers.dropout( - pool, - self._global_params.dropout_rate, - dropout_implementation='upscale_in_train') - - param_attr, bias_attr = init_fc_layer(class_dim, '_fc') - out = fluid.layers.fc( - pool, - class_dim, - name='_fc', - param_attr=param_attr, - bias_attr=bias_attr) - return out, pool - - def _drop_connect(self, inputs, prob, is_test): - if is_test: - return inputs - keep_prob = 1.0 - prob - random_tensor = keep_prob + fluid.layers.uniform_random_batch_size_like( - inputs, [-1, 1, 1, 1], min=0., max=1.) - binary_tensor = fluid.layers.floor(random_tensor) - output = inputs / keep_prob * binary_tensor - return output - - def _expand_conv_norm(self, inputs, block_args, is_test, name=None): - # Expansion phase - oup = block_args.input_filters * block_args.expand_ratio # number of output channels - - if block_args.expand_ratio != 1: - conv = self.conv_bn_layer( - inputs, - num_filters=oup, - filter_size=1, - bn_act=None, - bn_mom=self._bn_mom, - bn_eps=self._bn_eps, - padding_type=self.padding_type, - name=name, - conv_name=name + '_expand_conv', - bn_name='_bn0') - - return conv - - def _depthwise_conv_norm(self, inputs, block_args, is_test, name=None): - k = block_args.kernel_size - s = block_args.stride - if isinstance(s, list) or isinstance(s, tuple): - s = s[0] - oup = block_args.input_filters * block_args.expand_ratio # number of output channels - - conv = self.conv_bn_layer( - inputs, - num_filters=oup, - filter_size=k, - stride=s, - num_groups=oup, - bn_act=None, - padding_type=self.padding_type, - bn_mom=self._bn_mom, - bn_eps=self._bn_eps, - name=name, - use_cudnn=False, - conv_name=name + '_depthwise_conv', - bn_name='_bn1') - - return conv - - def _project_conv_norm(self, inputs, block_args, is_test, name=None): - final_oup = block_args.output_filters - conv = self.conv_bn_layer( - inputs, - num_filters=final_oup, - filter_size=1, - bn_act=None, - padding_type=self.padding_type, - bn_mom=self._bn_mom, - bn_eps=self._bn_eps, - name=name, - conv_name=name + '_project_conv', - bn_name='_bn2') - return conv - - def conv_bn_layer(self, - input, - filter_size, - num_filters, - stride=1, - num_groups=1, - padding_type="SAME", - conv_act=None, - bn_act='swish', - use_cudnn=True, - use_bn=True, - bn_mom=0.9, - bn_eps=1e-05, - use_bias=False, - name=None, - conv_name=None, - bn_name=None): - conv = conv2d( - input=input, - num_filters=num_filters, - filter_size=filter_size, - stride=stride, - groups=num_groups, - act=conv_act, - padding_type=padding_type, - use_cudnn=use_cudnn, - name=conv_name, - use_bias=use_bias) - - if use_bn == False: - return conv - else: - bn_name = name + bn_name - param_attr, bias_attr = init_batch_norm_layer(bn_name) - return fluid.layers.batch_norm( - input=conv, - act=bn_act, - momentum=bn_mom, - epsilon=bn_eps, - name=bn_name, - moving_mean_name=bn_name + '_mean', - moving_variance_name=bn_name + '_variance', - param_attr=param_attr, - bias_attr=bias_attr) - - def _conv_stem_norm(self, inputs, is_test): - out_channels = round_filters(32, self._global_params) - bn = self.conv_bn_layer( - inputs, - num_filters=out_channels, - filter_size=3, - stride=2, - bn_act=None, - bn_mom=self._bn_mom, - padding_type=self.padding_type, - bn_eps=self._bn_eps, - name='', - conv_name='_conv_stem', - bn_name='_bn0') - - return bn - - def mb_conv_block(self, - inputs, - block_args, - is_test=False, - drop_connect_rate=None, - name=None): - # Expansion and Depthwise Convolution - oup = block_args.input_filters * block_args.expand_ratio # number of output channels - has_se = self.use_se and (block_args.se_ratio is - not None) and (0 < block_args.se_ratio <= 1) - id_skip = block_args.id_skip # skip connection and drop connect - conv = inputs - if block_args.expand_ratio != 1: - conv = fluid.layers.swish( - self._expand_conv_norm(conv, block_args, is_test, name)) - - conv = fluid.layers.swish( - self._depthwise_conv_norm(conv, block_args, is_test, name)) - - # Squeeze and Excitation - if has_se: - num_squeezed_channels = max( - 1, int(block_args.input_filters * block_args.se_ratio)) - conv = self.se_block(conv, num_squeezed_channels, oup, name) - - conv = self._project_conv_norm(conv, block_args, is_test, name) - - # Skip connection and drop connect - input_filters, output_filters = block_args.input_filters, block_args.output_filters - if id_skip and block_args.stride == 1 and input_filters == output_filters: - if drop_connect_rate: - conv = self._drop_connect(conv, drop_connect_rate, is_test) - conv = fluid.layers.elementwise_add(conv, inputs) - - return conv - - def se_block(self, inputs, num_squeezed_channels, oup, name): - x_squeezed = fluid.layers.pool2d( - input=inputs, pool_type='avg', global_pooling=True, use_cudnn=False) - x_squeezed = conv2d( - x_squeezed, - num_filters=num_squeezed_channels, - filter_size=1, - use_bias=True, - padding_type=self.padding_type, - act='swish', - name=name + '_se_reduce') - x_squeezed = conv2d( - x_squeezed, - num_filters=oup, - filter_size=1, - use_bias=True, - padding_type=self.padding_type, - name=name + '_se_expand') - se_out = inputs * fluid.layers.sigmoid(x_squeezed) - return se_out - - def extract_features(self, inputs, is_test): - """ Returns output of the final convolution layer """ - - conv = fluid.layers.swish(self._conv_stem_norm(inputs, is_test=is_test)) - - block_args_copy = copy.deepcopy(self._blocks_args) - idx = 0 - block_size = 0 - for block_arg in block_args_copy: - block_arg = block_arg._replace( - input_filters=round_filters(block_arg.input_filters, - self._global_params), - output_filters=round_filters(block_arg.output_filters, - self._global_params), - num_repeat=round_repeats(block_arg.num_repeat, - self._global_params)) - block_size += 1 - for _ in range(block_arg.num_repeat - 1): - block_size += 1 - - for block_args in self._blocks_args: - - # Update block input and output filters based on depth multiplier. - block_args = block_args._replace( - input_filters=round_filters(block_args.input_filters, - self._global_params), - output_filters=round_filters(block_args.output_filters, - self._global_params), - num_repeat=round_repeats(block_args.num_repeat, - self._global_params)) - - # The first block needs to take care of stride and filter size increase. - drop_connect_rate = self._global_params.drop_connect_rate - if drop_connect_rate: - drop_connect_rate *= float(idx) / block_size - conv = self.mb_conv_block(conv, block_args, is_test, - drop_connect_rate, - '_blocks.' + str(idx) + '.') - - idx += 1 - if block_args.num_repeat > 1: - block_args = block_args._replace( - input_filters=block_args.output_filters, stride=1) - for _ in range(block_args.num_repeat - 1): - drop_connect_rate = self._global_params.drop_connect_rate - if drop_connect_rate: - drop_connect_rate *= float(idx) / block_size - conv = self.mb_conv_block(conv, block_args, is_test, - drop_connect_rate, - '_blocks.' + str(idx) + '.') - idx += 1 - - return conv - - def shortcut(self, input, data_residual): - return fluid.layers.elementwise_add(input, data_residual) - - -class BlockDecoder(object): - """ Block Decoder for readability, straight from the official TensorFlow repository """ - - @staticmethod - def _decode_block_string(block_string): - """ Gets a block through a string notation of arguments. """ - assert isinstance(block_string, str) - - ops = block_string.split('_') - options = {} - for op in ops: - splits = re.split(r'(\d.*)', op) - if len(splits) >= 2: - key, value = splits[:2] - options[key] = value - - # Check stride - assert (('s' in options and len(options['s']) == 1) or - (len(options['s']) == 2 and options['s'][0] == options['s'][1])) - - return BlockArgs( - kernel_size=int(options['k']), - num_repeat=int(options['r']), - input_filters=int(options['i']), - output_filters=int(options['o']), - expand_ratio=int(options['e']), - id_skip=('noskip' not in block_string), - se_ratio=float(options['se']) if 'se' in options else None, - stride=[int(options['s'][0])]) - - @staticmethod - def _encode_block_string(block): - """Encodes a block to a string.""" - args = [ - 'r%d' % block.num_repeat, - 'k%d' % block.kernel_size, - 's%d%d' % (block.strides[0], block.strides[1]), - 'e%s' % block.expand_ratio, - 'i%d' % block.input_filters, - 'o%d' % block.output_filters - ] - if 0 < block.se_ratio <= 1: - args.append('se%s' % block.se_ratio) - if block.id_skip is False: - args.append('noskip') - return '_'.join(args) - - @staticmethod - def decode(string_list): - """ - Decodes a list of string notations to specify blocks inside the network. - :param string_list: a list of strings, each string is a notation of block - :return: a list of BlockArgs namedtuples of block args - """ - assert isinstance(string_list, list) - blocks_args = [] - for block_string in string_list: - blocks_args.append(BlockDecoder._decode_block_string(block_string)) - return blocks_args - - @staticmethod - def encode(blocks_args): - """ - Encodes a list of BlockArgs to a list of strings. - :param blocks_args: a list of BlockArgs namedtuples of block args - :return: a list of strings, each string is a notation of block - """ - block_strings = [] - for block in blocks_args: - block_strings.append(BlockDecoder._encode_block_string(block)) - return block_strings - - -def EfficientNetB0_small(is_test=False, - padding_type='SAME', - override_params=None, - use_se=False): - model = EfficientNet( - name='b0', - is_test=is_test, - padding_type=padding_type, - override_params=override_params, - use_se=use_se) - return model - - -def EfficientNetB0(is_test=False, - padding_type='SAME', - override_params=None, - use_se=True): - model = EfficientNet( - name='b0', - is_test=is_test, - padding_type=padding_type, - override_params=override_params, - use_se=use_se) - return model - - -def EfficientNetB1(is_test=False, - padding_type='SAME', - override_params=None, - use_se=True): - model = EfficientNet( - name='b1', - is_test=is_test, - padding_type=padding_type, - override_params=override_params, - use_se=use_se) - return model - - -def EfficientNetB2(is_test=False, - padding_type='SAME', - override_params=None, - use_se=True): - model = EfficientNet( - name='b2', - is_test=is_test, - padding_type=padding_type, - override_params=override_params, - use_se=use_se) - return model - - -def EfficientNetB3(is_test=False, - padding_type='SAME', - override_params=None, - use_se=True): - model = EfficientNet( - name='b3', - is_test=is_test, - padding_type=padding_type, - override_params=override_params, - use_se=use_se) - return model - - -def EfficientNetB4(is_test=False, - padding_type='SAME', - override_params=None, - use_se=True): - model = EfficientNet( - name='b4', - is_test=is_test, - padding_type=padding_type, - override_params=override_params, - use_se=use_se) - return model - - -def EfficientNetB5(is_test=False, - padding_type='SAME', - override_params=None, - use_se=True): - model = EfficientNet( - name='b5', - is_test=is_test, - padding_type=padding_type, - override_params=override_params, - use_se=use_se) - return model - - -def EfficientNetB6(is_test=False, - padding_type='SAME', - override_params=None, - use_se=True): - model = EfficientNet( - name='b6', - is_test=is_test, - padding_type=padding_type, - override_params=override_params, - use_se=use_se) - return model - - -def EfficientNetB7(is_test=False, - padding_type='SAME', - override_params=None, - use_se=True): - model = EfficientNet( - name='b7', - is_test=is_test, - padding_type=padding_type, - override_params=override_params, - use_se=use_se) - return model diff --git a/hub_module/modules/image/classification/efficientnetb6_imagenet/label_list.txt b/hub_module/modules/image/classification/efficientnetb6_imagenet/label_list.txt deleted file mode 100644 index a509c007..00000000 --- a/hub_module/modules/image/classification/efficientnetb6_imagenet/label_list.txt +++ /dev/null @@ -1,1000 +0,0 @@ -tench, Tinca tinca -goldfish, Carassius auratus -great white shark, white shark, man-eater, man-eating shark, Carcharodon carcharias -tiger shark, Galeocerdo cuvieri -hammerhead, hammerhead shark -electric ray, crampfish, numbfish, torpedo -stingray -cock -hen -ostrich, Struthio camelus -brambling, Fringilla montifringilla -goldfinch, Carduelis carduelis -house finch, linnet, Carpodacus mexicanus -junco, snowbird -indigo bunting, indigo finch, indigo bird, Passerina cyanea -robin, American robin, Turdus migratorius -bulbul -jay -magpie -chickadee -water ouzel, dipper -kite -bald eagle, American eagle, Haliaeetus leucocephalus -vulture -great grey owl, great gray owl, Strix nebulosa -European fire salamander, Salamandra salamandra -common newt, Triturus vulgaris -eft -spotted salamander, Ambystoma maculatum -axolotl, mud puppy, Ambystoma mexicanum -bullfrog, Rana catesbeiana -tree frog, tree-frog -tailed frog, bell toad, ribbed toad, tailed toad, Ascaphus trui -loggerhead, loggerhead turtle, Caretta caretta -leatherback turtle, leatherback, leathery turtle, Dermochelys coriacea -mud turtle -terrapin -box turtle, box tortoise -banded gecko -common iguana, iguana, Iguana iguana -American chameleon, anole, Anolis carolinensis -whiptail, whiptail lizard -agama -frilled lizard, Chlamydosaurus kingi -alligator lizard -Gila monster, Heloderma suspectum -green lizard, Lacerta viridis -African chameleon, Chamaeleo chamaeleon -Komodo dragon, Komodo lizard, dragon lizard, giant lizard, Varanus komodoensis -African crocodile, Nile crocodile, Crocodylus niloticus -American alligator, Alligator mississipiensis -triceratops -thunder snake, worm snake, Carphophis amoenus -ringneck snake, ring-necked snake, ring snake -hognose snake, puff adder, sand viper -green snake, grass snake -king snake, kingsnake -garter snake, grass snake -water snake -vine snake -night snake, Hypsiglena torquata -boa constrictor, Constrictor constrictor -rock python, rock snake, Python sebae -Indian cobra, Naja naja -green mamba -sea snake -horned viper, cerastes, sand viper, horned asp, Cerastes cornutus -diamondback, diamondback rattlesnake, Crotalus adamanteus -sidewinder, horned rattlesnake, Crotalus cerastes -trilobite -harvestman, daddy longlegs, Phalangium opilio -scorpion -black and gold garden spider, Argiope aurantia -barn spider, Araneus cavaticus -garden spider, Aranea diademata -black widow, Latrodectus mactans -tarantula -wolf spider, hunting spider -tick -centipede -black grouse -ptarmigan -ruffed grouse, partridge, Bonasa umbellus -prairie chicken, prairie grouse, prairie fowl -peacock -quail -partridge -African grey, African gray, Psittacus erithacus -macaw -sulphur-crested cockatoo, Kakatoe galerita, Cacatua galerita -lorikeet -coucal -bee eater -hornbill -hummingbird -jacamar -toucan -drake -red-breasted merganser, Mergus serrator -goose -black swan, Cygnus atratus -tusker -echidna, spiny anteater, anteater -platypus, duckbill, duckbilled platypus, duck-billed platypus, Ornithorhynchus anatinus -wallaby, brush kangaroo -koala, koala bear, kangaroo bear, native bear, Phascolarctos cinereus -wombat -jellyfish -sea anemone, anemone -brain coral -flatworm, platyhelminth -nematode, nematode worm, roundworm -conch -snail -slug -sea slug, nudibranch -chiton, coat-of-mail shell, sea cradle, polyplacophore -chambered nautilus, pearly nautilus, nautilus -Dungeness crab, Cancer magister -rock crab, Cancer irroratus -fiddler crab -king crab, Alaska crab, Alaskan king crab, Alaska king crab, Paralithodes camtschatica -American lobster, Northern lobster, Maine lobster, Homarus americanus -spiny lobster, langouste, rock lobster, crawfish, crayfish, sea crawfish -crayfish, crawfish, crawdad, crawdaddy -hermit crab -isopod -white stork, Ciconia ciconia -black stork, Ciconia nigra -spoonbill -flamingo -little blue heron, Egretta caerulea -American egret, great white heron, Egretta albus -bittern -crane -limpkin, Aramus pictus -European gallinule, Porphyrio porphyrio -American coot, marsh hen, mud hen, water hen, Fulica americana -bustard -ruddy turnstone, Arenaria interpres -red-backed sandpiper, dunlin, Erolia alpina -redshank, Tringa totanus -dowitcher -oystercatcher, oyster catcher -pelican -king penguin, Aptenodytes patagonica -albatross, mollymawk -grey whale, gray whale, devilfish, Eschrichtius gibbosus, Eschrichtius robustus -killer whale, killer, orca, grampus, sea wolf, Orcinus orca -dugong, Dugong dugon -sea lion -Chihuahua -Japanese spaniel -Maltese dog, Maltese terrier, Maltese -Pekinese, Pekingese, Peke -Shih-Tzu -Blenheim spaniel -papillon -toy terrier -Rhodesian ridgeback -Afghan hound, Afghan -basset, basset hound -beagle -bloodhound, sleuthhound -bluetick -black-and-tan coonhound -Walker hound, Walker foxhound -English foxhound -redbone -borzoi, Russian wolfhound -Irish wolfhound -Italian greyhound -whippet -Ibizan hound, Ibizan Podenco -Norwegian elkhound, elkhound -otterhound, otter hound -Saluki, gazelle hound -Scottish deerhound, deerhound -Weimaraner -Staffordshire bullterrier, Staffordshire bull terrier -American Staffordshire terrier, Staffordshire terrier, American pit bull terrier, pit bull terrier -Bedlington terrier -Border terrier -Kerry blue terrier -Irish terrier -Norfolk terrier -Norwich terrier -Yorkshire terrier -wire-haired fox terrier -Lakeland terrier -Sealyham terrier, Sealyham -Airedale, Airedale terrier -cairn, cairn terrier -Australian terrier -Dandie Dinmont, Dandie Dinmont terrier -Boston bull, Boston terrier -miniature schnauzer -giant schnauzer -standard schnauzer -Scotch terrier, Scottish terrier, Scottie -Tibetan terrier, chrysanthemum dog -silky terrier, Sydney silky -soft-coated wheaten terrier -West Highland white terrier -Lhasa, Lhasa apso -flat-coated retriever -curly-coated retriever -golden retriever -Labrador retriever -Chesapeake Bay retriever -German short-haired pointer -vizsla, Hungarian pointer -English setter -Irish setter, red setter -Gordon setter -Brittany spaniel -clumber, clumber spaniel -English springer, English springer spaniel -Welsh springer spaniel -cocker spaniel, English cocker spaniel, cocker -Sussex spaniel -Irish water spaniel -kuvasz -schipperke -groenendael -malinois -briard -kelpie -komondor -Old English sheepdog, bobtail -Shetland sheepdog, Shetland sheep dog, Shetland -collie -Border collie -Bouvier des Flandres, Bouviers des Flandres -Rottweiler -German shepherd, German shepherd dog, German police dog, alsatian -Doberman, Doberman pinscher -miniature pinscher -Greater Swiss Mountain dog -Bernese mountain dog -Appenzeller -EntleBucher -boxer -bull mastiff -Tibetan mastiff -French bulldog -Great Dane -Saint Bernard, St Bernard -Eskimo dog, husky -malamute, malemute, Alaskan malamute -Siberian husky -dalmatian, coach dog, carriage dog -affenpinscher, monkey pinscher, monkey dog -basenji -pug, pug-dog -Leonberg -Newfoundland, Newfoundland dog -Great Pyrenees -Samoyed, Samoyede -Pomeranian -chow, chow chow -keeshond -Brabancon griffon -Pembroke, Pembroke Welsh corgi -Cardigan, Cardigan Welsh corgi -toy poodle -miniature poodle -standard poodle -Mexican hairless -timber wolf, grey wolf, gray wolf, Canis lupus -white wolf, Arctic wolf, Canis lupus tundrarum -red wolf, maned wolf, Canis rufus, Canis niger -coyote, prairie wolf, brush wolf, Canis latrans -dingo, warrigal, warragal, Canis dingo -dhole, Cuon alpinus -African hunting dog, hyena dog, Cape hunting dog, Lycaon pictus -hyena, hyaena -red fox, Vulpes vulpes -kit fox, Vulpes macrotis -Arctic fox, white fox, Alopex lagopus -grey fox, gray fox, Urocyon cinereoargenteus -tabby, tabby cat -tiger cat -Persian cat -Siamese cat, Siamese -Egyptian cat -cougar, puma, catamount, mountain lion, painter, panther, Felis concolor -lynx, catamount -leopard, Panthera pardus -snow leopard, ounce, Panthera uncia -jaguar, panther, Panthera onca, Felis onca -lion, king of beasts, Panthera leo -tiger, Panthera tigris -cheetah, chetah, Acinonyx jubatus -brown bear, bruin, Ursus arctos -American black bear, black bear, Ursus americanus, Euarctos americanus -ice bear, polar bear, Ursus Maritimus, Thalarctos maritimus -sloth bear, Melursus ursinus, Ursus ursinus -mongoose -meerkat, mierkat -tiger beetle -ladybug, ladybeetle, lady beetle, ladybird, ladybird beetle -ground beetle, carabid beetle -long-horned beetle, longicorn, longicorn beetle -leaf beetle, chrysomelid -dung beetle -rhinoceros beetle -weevil -fly -bee -ant, emmet, pismire -grasshopper, hopper -cricket -walking stick, walkingstick, stick insect -cockroach, roach -mantis, mantid -cicada, cicala -leafhopper -lacewing, lacewing fly -dragonfly, darning needle, devil's darning needle, sewing needle, snake feeder, snake doctor, mosquito hawk, skeeter hawk -damselfly -admiral -ringlet, ringlet butterfly -monarch, monarch butterfly, milkweed butterfly, Danaus plexippus -cabbage butterfly -sulphur butterfly, sulfur butterfly -lycaenid, lycaenid butterfly -starfish, sea star -sea urchin -sea cucumber, holothurian -wood rabbit, cottontail, cottontail rabbit -hare -Angora, Angora rabbit -hamster -porcupine, hedgehog -fox squirrel, eastern fox squirrel, Sciurus niger -marmot -beaver -guinea pig, Cavia cobaya -sorrel -zebra -hog, pig, grunter, squealer, Sus scrofa -wild boar, boar, Sus scrofa -warthog -hippopotamus, hippo, river horse, Hippopotamus amphibius -ox -water buffalo, water ox, Asiatic buffalo, Bubalus bubalis -bison -ram, tup -bighorn, bighorn sheep, cimarron, Rocky Mountain bighorn, Rocky Mountain sheep, Ovis canadensis -ibex, Capra ibex -hartebeest -impala, Aepyceros melampus -gazelle -Arabian camel, dromedary, Camelus dromedarius -llama -weasel -mink -polecat, fitch, foulmart, foumart, Mustela putorius -black-footed ferret, ferret, Mustela nigripes -otter -skunk, polecat, wood pussy -badger -armadillo -three-toed sloth, ai, Bradypus tridactylus -orangutan, orang, orangutang, Pongo pygmaeus -gorilla, Gorilla gorilla -chimpanzee, chimp, Pan troglodytes -gibbon, Hylobates lar -siamang, Hylobates syndactylus, Symphalangus syndactylus -guenon, guenon monkey -patas, hussar monkey, Erythrocebus patas -baboon -macaque -langur -colobus, colobus monkey -proboscis monkey, Nasalis larvatus -marmoset -capuchin, ringtail, Cebus capucinus -howler monkey, howler -titi, titi monkey -spider monkey, Ateles geoffroyi -squirrel monkey, Saimiri sciureus -Madagascar cat, ring-tailed lemur, Lemur catta -indri, indris, Indri indri, Indri brevicaudatus -Indian elephant, Elephas maximus -African elephant, Loxodonta africana -lesser panda, red panda, panda, bear cat, cat bear, Ailurus fulgens -giant panda, panda, panda bear, coon bear, Ailuropoda melanoleuca -barracouta, snoek -eel -coho, cohoe, coho salmon, blue jack, silver salmon, Oncorhynchus kisutch -rock beauty, Holocanthus tricolor -anemone fish -sturgeon -gar, garfish, garpike, billfish, Lepisosteus osseus -lionfish -puffer, pufferfish, blowfish, globefish -abacus -abaya -academic gown, academic robe, judge's robe -accordion, piano accordion, squeeze box -acoustic guitar -aircraft carrier, carrier, flattop, attack aircraft carrier -airliner -airship, dirigible -altar -ambulance -amphibian, amphibious vehicle -analog clock -apiary, bee house -apron -ashcan, trash can, garbage can, wastebin, ash bin, ash-bin, ashbin, dustbin, trash barrel, trash bin -assault rifle, assault gun -backpack, back pack, knapsack, packsack, rucksack, haversack -bakery, bakeshop, bakehouse -balance beam, beam -balloon -ballpoint, ballpoint pen, ballpen, Biro -Band Aid -banjo -bannister, banister, balustrade, balusters, handrail -barbell -barber chair -barbershop -barn -barometer -barrel, cask -barrow, garden cart, lawn cart, wheelbarrow -baseball -basketball -bassinet -bassoon -bathing cap, swimming cap -bath towel -bathtub, bathing tub, bath, tub -beach wagon, station wagon, wagon, estate car, beach waggon, station waggon, waggon -beacon, lighthouse, beacon light, pharos -beaker -bearskin, busby, shako -beer bottle -beer glass -bell cote, bell cot -bib -bicycle-built-for-two, tandem bicycle, tandem -bikini, two-piece -binder, ring-binder -binoculars, field glasses, opera glasses -birdhouse -boathouse -bobsled, bobsleigh, bob -bolo tie, bolo, bola tie, bola -bonnet, poke bonnet -bookcase -bookshop, bookstore, bookstall -bottlecap -bow -bow tie, bow-tie, bowtie -brass, memorial tablet, plaque -brassiere, bra, bandeau -breakwater, groin, groyne, mole, bulwark, seawall, jetty -breastplate, aegis, egis -broom -bucket, pail -buckle -bulletproof vest -bullet train, bullet -butcher shop, meat market -cab, hack, taxi, taxicab -caldron, cauldron -candle, taper, wax light -cannon -canoe -can opener, tin opener -cardigan -car mirror -carousel, carrousel, merry-go-round, roundabout, whirligig -carpenter's kit, tool kit -carton -car wheel -cash machine, cash dispenser, automated teller machine, automatic teller machine, automated teller, automatic teller, ATM -cassette -cassette player -castle -catamaran -CD player -cello, violoncello -cellular telephone, cellular phone, cellphone, cell, mobile phone -chain -chainlink fence -chain mail, ring mail, mail, chain armor, chain armour, ring armor, ring armour -chain saw, chainsaw -chest -chiffonier, commode -chime, bell, gong -china cabinet, china closet -Christmas stocking -church, church building -cinema, movie theater, movie theatre, movie house, picture palace -cleaver, meat cleaver, chopper -cliff dwelling -cloak -clog, geta, patten, sabot -cocktail shaker -coffee mug -coffeepot -coil, spiral, volute, whorl, helix -combination lock -computer keyboard, keypad -confectionery, confectionary, candy store -container ship, containership, container vessel -convertible -corkscrew, bottle screw -cornet, horn, trumpet, trump -cowboy boot -cowboy hat, ten-gallon hat -cradle -crane -crash helmet -crate -crib, cot -Crock Pot -croquet ball -crutch -cuirass -dam, dike, dyke -desk -desktop computer -dial telephone, dial phone -diaper, nappy, napkin -digital clock -digital watch -dining table, board -dishrag, dishcloth -dishwasher, dish washer, dishwashing machine -disk brake, disc brake -dock, dockage, docking facility -dogsled, dog sled, dog sleigh -dome -doormat, welcome mat -drilling platform, offshore rig -drum, membranophone, tympan -drumstick -dumbbell -Dutch oven -electric fan, blower -electric guitar -electric locomotive -entertainment center -envelope -espresso maker -face powder -feather boa, boa -file, file cabinet, filing cabinet -fireboat -fire engine, fire truck -fire screen, fireguard -flagpole, flagstaff -flute, transverse flute -folding chair -football helmet -forklift -fountain -fountain pen -four-poster -freight car -French horn, horn -frying pan, frypan, skillet -fur coat -garbage truck, dustcart -gasmask, respirator, gas helmet -gas pump, gasoline pump, petrol pump, island dispenser -goblet -go-kart -golf ball -golfcart, golf cart -gondola -gong, tam-tam -gown -grand piano, grand -greenhouse, nursery, glasshouse -grille, radiator grille -grocery store, grocery, food market, market -guillotine -hair slide -hair spray -half track -hammer -hamper -hand blower, blow dryer, blow drier, hair dryer, hair drier -hand-held computer, hand-held microcomputer -handkerchief, hankie, hanky, hankey -hard disc, hard disk, fixed disk -harmonica, mouth organ, harp, mouth harp -harp -harvester, reaper -hatchet -holster -home theater, home theatre -honeycomb -hook, claw -hoopskirt, crinoline -horizontal bar, high bar -horse cart, horse-cart -hourglass -iPod -iron, smoothing iron -jack-o'-lantern -jean, blue jean, denim -jeep, landrover -jersey, T-shirt, tee shirt -jigsaw puzzle -jinrikisha, ricksha, rickshaw -joystick -kimono -knee pad -knot -lab coat, laboratory coat -ladle -lampshade, lamp shade -laptop, laptop computer -lawn mower, mower -lens cap, lens cover -letter opener, paper knife, paperknife -library -lifeboat -lighter, light, igniter, ignitor -limousine, limo -liner, ocean liner -lipstick, lip rouge -Loafer -lotion -loudspeaker, speaker, speaker unit, loudspeaker system, speaker system -loupe, jeweler's loupe -lumbermill, sawmill -magnetic compass -mailbag, postbag -mailbox, letter box -maillot -maillot, tank suit -manhole cover -maraca -marimba, xylophone -mask -matchstick -maypole -maze, labyrinth -measuring cup -medicine chest, medicine cabinet -megalith, megalithic structure -microphone, mike -microwave, microwave oven -military uniform -milk can -minibus -miniskirt, mini -minivan -missile -mitten -mixing bowl -mobile home, manufactured home -Model T -modem -monastery -monitor -moped -mortar -mortarboard -mosque -mosquito net -motor scooter, scooter -mountain bike, all-terrain bike, off-roader -mountain tent -mouse, computer mouse -mousetrap -moving van -muzzle -nail -neck brace -necklace -nipple -notebook, notebook computer -obelisk -oboe, hautboy, hautbois -ocarina, sweet potato -odometer, hodometer, mileometer, milometer -oil filter -organ, pipe organ -oscilloscope, scope, cathode-ray oscilloscope, CRO -overskirt -oxcart -oxygen mask -packet -paddle, boat paddle -paddlewheel, paddle wheel -padlock -paintbrush -pajama, pyjama, pj's, jammies -palace -panpipe, pandean pipe, syrinx -paper towel -parachute, chute -parallel bars, bars -park bench -parking meter -passenger car, coach, carriage -patio, terrace -pay-phone, pay-station -pedestal, plinth, footstall -pencil box, pencil case -pencil sharpener -perfume, essence -Petri dish -photocopier -pick, plectrum, plectron -pickelhaube -picket fence, paling -pickup, pickup truck -pier -piggy bank, penny bank -pill bottle -pillow -ping-pong ball -pinwheel -pirate, pirate ship -pitcher, ewer -plane, carpenter's plane, woodworking plane -planetarium -plastic bag -plate rack -plow, plough -plunger, plumber's helper -Polaroid camera, Polaroid Land camera -pole -police van, police wagon, paddy wagon, patrol wagon, wagon, black Maria -poncho -pool table, billiard table, snooker table -pop bottle, soda bottle -pot, flowerpot -potter's wheel -power drill -prayer rug, prayer mat -printer -prison, prison house -projectile, missile -projector -puck, hockey puck -punching bag, punch bag, punching ball, punchball -purse -quill, quill pen -quilt, comforter, comfort, puff -racer, race car, racing car -racket, racquet -radiator -radio, wireless -radio telescope, radio reflector -rain barrel -recreational vehicle, RV, R.V. -reel -reflex camera -refrigerator, icebox -remote control, remote -restaurant, eating house, eating place, eatery -revolver, six-gun, six-shooter -rifle -rocking chair, rocker -rotisserie -rubber eraser, rubber, pencil eraser -rugby ball -rule, ruler -running shoe -safe -safety pin -saltshaker, salt shaker -sandal -sarong -sax, saxophone -scabbard -scale, weighing machine -school bus -schooner -scoreboard -screen, CRT screen -screw -screwdriver -seat belt, seatbelt -sewing machine -shield, buckler -shoe shop, shoe-shop, shoe store -shoji -shopping basket -shopping cart -shovel -shower cap -shower curtain -ski -ski mask -sleeping bag -slide rule, slipstick -sliding door -slot, one-armed bandit -snorkel -snowmobile -snowplow, snowplough -soap dispenser -soccer ball -sock -solar dish, solar collector, solar furnace -sombrero -soup bowl -space bar -space heater -space shuttle -spatula -speedboat -spider web, spider's web -spindle -sports car, sport car -spotlight, spot -stage -steam locomotive -steel arch bridge -steel drum -stethoscope -stole -stone wall -stopwatch, stop watch -stove -strainer -streetcar, tram, tramcar, trolley, trolley car -stretcher -studio couch, day bed -stupa, tope -submarine, pigboat, sub, U-boat -suit, suit of clothes -sundial -sunglass -sunglasses, dark glasses, shades -sunscreen, sunblock, sun blocker -suspension bridge -swab, swob, mop -sweatshirt -swimming trunks, bathing trunks -swing -switch, electric switch, electrical switch -syringe -table lamp -tank, army tank, armored combat vehicle, armoured combat vehicle -tape player -teapot -teddy, teddy bear -television, television system -tennis ball -thatch, thatched roof -theater curtain, theatre curtain -thimble -thresher, thrasher, threshing machine -throne -tile roof -toaster -tobacco shop, tobacconist shop, tobacconist -toilet seat -torch -totem pole -tow truck, tow car, wrecker -toyshop -tractor -trailer truck, tractor trailer, trucking rig, rig, articulated lorry, semi -tray -trench coat -tricycle, trike, velocipede -trimaran -tripod -triumphal arch -trolleybus, trolley coach, trackless trolley -trombone -tub, vat -turnstile -typewriter keyboard -umbrella -unicycle, monocycle -upright, upright piano -vacuum, vacuum cleaner -vase -vault -velvet -vending machine -vestment -viaduct -violin, fiddle -volleyball -waffle iron -wall clock -wallet, billfold, notecase, pocketbook -wardrobe, closet, press -warplane, military plane -washbasin, handbasin, washbowl, lavabo, wash-hand basin -washer, automatic washer, washing machine -water bottle -water jug -water tower -whiskey jug -whistle -wig -window screen -window shade -Windsor tie -wine bottle -wing -wok -wooden spoon -wool, woolen, woollen -worm fence, snake fence, snake-rail fence, Virginia fence -wreck -yawl -yurt -web site, website, internet site, site -comic book -crossword puzzle, crossword -street sign -traffic light, traffic signal, stoplight -book jacket, dust cover, dust jacket, dust wrapper -menu -plate -guacamole -consomme -hot pot, hotpot -trifle -ice cream, icecream -ice lolly, lolly, lollipop, popsicle -French loaf -bagel, beigel -pretzel -cheeseburger -hotdog, hot dog, red hot -mashed potato -head cabbage -broccoli -cauliflower -zucchini, courgette -spaghetti squash -acorn squash -butternut squash -cucumber, cuke -artichoke, globe artichoke -bell pepper -cardoon -mushroom -Granny Smith -strawberry -orange -lemon -fig -pineapple, ananas -banana -jackfruit, jak, jack -custard apple -pomegranate -hay -carbonara -chocolate sauce, chocolate syrup -dough -meat loaf, meatloaf -pizza, pizza pie -potpie -burrito -red wine -espresso -cup -eggnog -alp -bubble -cliff, drop, drop-off -coral reef -geyser -lakeside, lakeshore -promontory, headland, head, foreland -sandbar, sand bar -seashore, coast, seacoast, sea-coast -valley, vale -volcano -ballplayer, baseball player -groom, bridegroom -scuba diver -rapeseed -daisy -yellow lady's slipper, yellow lady-slipper, Cypripedium calceolus, Cypripedium parviflorum -corn -acorn -hip, rose hip, rosehip -buckeye, horse chestnut, conker -coral fungus -agaric -gyromitra -stinkhorn, carrion fungus -earthstar -hen-of-the-woods, hen of the woods, Polyporus frondosus, Grifola frondosa -bolete -ear, spike, capitulum -toilet tissue, toilet paper, bathroom tissue diff --git a/hub_module/modules/image/classification/efficientnetb6_imagenet/layers.py b/hub_module/modules/image/classification/efficientnetb6_imagenet/layers.py deleted file mode 100644 index 00f792b5..00000000 --- a/hub_module/modules/image/classification/efficientnetb6_imagenet/layers.py +++ /dev/null @@ -1,248 +0,0 @@ -# -*- coding:utf-8 -*- -# copyright (c) 2020 PaddlePaddle Authors. All Rights Reserve. -# -# Licensed under the Apache License, Version 2.0 (the "License"); -# you may not use this file except in compliance with the License. -# You may obtain a copy of the License at -# -# http://www.apache.org/licenses/LICENSE-2.0 -# -# Unless required by applicable law or agreed to in writing, software -# distributed under the License is distributed on an "AS IS" BASIS, -# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. -# See the License for the specific language governing permissions and -# limitations under the License. - -from __future__ import absolute_import -from __future__ import division -from __future__ import print_function - -import math -import warnings - -import paddle.fluid as fluid - - -def initial_type(name, - input, - op_type, - fan_out, - init="google", - use_bias=False, - filter_size=0, - stddev=0.02): - if init == "kaiming": - if op_type == 'conv': - fan_in = input.shape[1] * filter_size * filter_size - elif op_type == 'deconv': - fan_in = fan_out * filter_size * filter_size - else: - if len(input.shape) > 2: - fan_in = input.shape[1] * input.shape[2] * input.shape[3] - else: - fan_in = input.shape[1] - bound = 1 / math.sqrt(fan_in) - param_attr = fluid.ParamAttr( - name=name + "_weights", - initializer=fluid.initializer.Uniform(low=-bound, high=bound)) - if use_bias == True: - bias_attr = fluid.ParamAttr( - name=name + '_offset', - initializer=fluid.initializer.Uniform(low=-bound, high=bound)) - else: - bias_attr = False - elif init == 'google': - n = filter_size * filter_size * fan_out - param_attr = fluid.ParamAttr( - name=name + "_weights", - initializer=fluid.initializer.NormalInitializer( - loc=0.0, scale=math.sqrt(2.0 / n))) - if use_bias == True: - bias_attr = fluid.ParamAttr( - name=name + "_offset", - initializer=fluid.initializer.Constant(0.0)) - else: - bias_attr = False - - else: - param_attr = fluid.ParamAttr( - name=name + "_weights", - initializer=fluid.initializer.NormalInitializer( - loc=0.0, scale=stddev)) - if use_bias == True: - bias_attr = fluid.ParamAttr( - name=name + "_offset", - initializer=fluid.initializer.Constant(0.0)) - else: - bias_attr = False - return param_attr, bias_attr - - -def cal_padding(img_size, stride, filter_size, dilation=1): - """Calculate padding size.""" - if img_size % stride == 0: - out_size = max(filter_size - stride, 0) - else: - out_size = max(filter_size - (img_size % stride), 0) - return out_size // 2, out_size - out_size // 2 - - -def init_batch_norm_layer(name="batch_norm"): - param_attr = fluid.ParamAttr( - name=name + '_scale', initializer=fluid.initializer.Constant(1.0)) - bias_attr = fluid.ParamAttr( - name=name + '_offset', - initializer=fluid.initializer.Constant(value=0.0)) - return param_attr, bias_attr - - -def init_fc_layer(fout, name='fc'): - n = fout # fan-out - init_range = 1.0 / math.sqrt(n) - - param_attr = fluid.ParamAttr( - name=name + '_weights', - initializer=fluid.initializer.UniformInitializer( - low=-init_range, high=init_range)) - bias_attr = fluid.ParamAttr( - name=name + '_offset', - initializer=fluid.initializer.Constant(value=0.0)) - return param_attr, bias_attr - - -def norm_layer(input, norm_type='batch_norm', name=None): - if norm_type == 'batch_norm': - param_attr = fluid.ParamAttr( - name=name + '_weights', initializer=fluid.initializer.Constant(1.0)) - bias_attr = fluid.ParamAttr( - name=name + '_offset', - initializer=fluid.initializer.Constant(value=0.0)) - return fluid.layers.batch_norm( - input, - param_attr=param_attr, - bias_attr=bias_attr, - moving_mean_name=name + '_mean', - moving_variance_name=name + '_variance') - - elif norm_type == 'instance_norm': - helper = fluid.layer_helper.LayerHelper("instance_norm", **locals()) - dtype = helper.input_dtype() - epsilon = 1e-5 - mean = fluid.layers.reduce_mean(input, dim=[2, 3], keep_dim=True) - var = fluid.layers.reduce_mean( - fluid.layers.square(input - mean), dim=[2, 3], keep_dim=True) - if name is not None: - scale_name = name + "_scale" - offset_name = name + "_offset" - scale_param = fluid.ParamAttr( - name=scale_name, - initializer=fluid.initializer.Constant(1.0), - trainable=True) - offset_param = fluid.ParamAttr( - name=offset_name, - initializer=fluid.initializer.Constant(0.0), - trainable=True) - scale = helper.create_parameter( - attr=scale_param, shape=input.shape[1:2], dtype=dtype) - offset = helper.create_parameter( - attr=offset_param, shape=input.shape[1:2], dtype=dtype) - - tmp = fluid.layers.elementwise_mul(x=(input - mean), y=scale, axis=1) - tmp = tmp / fluid.layers.sqrt(var + epsilon) - tmp = fluid.layers.elementwise_add(tmp, offset, axis=1) - return tmp - else: - raise NotImplementedError("norm tyoe: [%s] is not support" % norm_type) - - -def conv2d(input, - num_filters=64, - filter_size=7, - stride=1, - stddev=0.02, - padding=0, - groups=None, - name="conv2d", - norm=None, - act=None, - relufactor=0.0, - use_bias=False, - padding_type=None, - initial="normal", - use_cudnn=True): - - if padding != 0 and padding_type != None: - warnings.warn( - 'padding value and padding type are set in the same time, and the final padding width and padding height are computed by padding_type' - ) - - param_attr, bias_attr = initial_type( - name=name, - input=input, - op_type='conv', - fan_out=num_filters, - init=initial, - use_bias=use_bias, - filter_size=filter_size, - stddev=stddev) - - def get_padding(filter_size, stride=1, dilation=1): - padding = ((stride - 1) + dilation * (filter_size - 1)) // 2 - return padding - - need_crop = False - if padding_type == "SAME": - top_padding, bottom_padding = cal_padding(input.shape[2], stride, - filter_size) - left_padding, right_padding = cal_padding(input.shape[2], stride, - filter_size) - height_padding = bottom_padding - width_padding = right_padding - if top_padding != bottom_padding or left_padding != right_padding: - height_padding = top_padding + stride - width_padding = left_padding + stride - need_crop = True - padding = [height_padding, width_padding] - elif padding_type == "VALID": - height_padding = 0 - width_padding = 0 - padding = [height_padding, width_padding] - elif padding_type == "DYNAMIC": - padding = get_padding(filter_size, stride) - else: - padding = padding - - conv = fluid.layers.conv2d( - input, - num_filters, - filter_size, - groups=groups, - name=name, - stride=stride, - padding=padding, - use_cudnn=use_cudnn, - param_attr=param_attr, - bias_attr=bias_attr) - - if need_crop: - conv = conv[:, :, 1:, 1:] - - if norm is not None: - conv = norm_layer(input=conv, norm_type=norm, name=name + "_norm") - if act == 'relu': - conv = fluid.layers.relu(conv, name=name + '_relu') - elif act == 'leaky_relu': - conv = fluid.layers.leaky_relu( - conv, alpha=relufactor, name=name + '_leaky_relu') - elif act == 'tanh': - conv = fluid.layers.tanh(conv, name=name + '_tanh') - elif act == 'sigmoid': - conv = fluid.layers.sigmoid(conv, name=name + '_sigmoid') - elif act == 'swish': - conv = fluid.layers.swish(conv, name=name + '_swish') - elif act == None: - conv = conv - else: - raise NotImplementedError("activation: [%s] is not support" % act) - - return conv diff --git a/hub_module/modules/image/classification/efficientnetb6_imagenet/module.py b/hub_module/modules/image/classification/efficientnetb6_imagenet/module.py index af1427f6..fd3a4bd2 100644 --- a/hub_module/modules/image/classification/efficientnetb6_imagenet/module.py +++ b/hub_module/modules/image/classification/efficientnetb6_imagenet/module.py @@ -1,4 +1,3 @@ -# -*- coding:utf-8 -*- # copyright (c) 2020 PaddlePaddle Authors. All Rights Reserve. # # Licensed under the Apache License, Version 2.0 (the "License"); @@ -13,312 +12,747 @@ # See the License for the specific language governing permissions and # limitations under the License. -from __future__ import absolute_import -from __future__ import division - -import ast -import argparse import os - -import numpy as np -import paddle.fluid as fluid -import paddlehub as hub -from paddle.fluid.core import PaddleTensor, AnalysisConfig, create_paddle_predictor -from paddlehub.module.module import moduleinfo, runnable, serving -from paddlehub.common.paddle_helper import add_vars_prefix - -from efficientnetb6_imagenet.processor import postprocess, base64_to_cv2 -from efficientnetb6_imagenet.data_feed import reader -from efficientnetb6_imagenet.efficientnet import EfficientNetB6 - - -@moduleinfo( - name="efficientnetb6_imagenet", - type="CV/image_classification", - author="paddlepaddle", - author_email="paddle-dev@baidu.com", - summary= - "EfficientNetB6 is a image classfication model, this module is trained with imagenet datasets.", - version="1.1.0") -class EfficientNetB6ImageNet(hub.Module): - def _initialize(self): - self.default_pretrained_model_path = os.path.join( - self.directory, "efficientnetb6_imagenet_infer_model") - label_file = os.path.join(self.directory, "label_list.txt") - with open(label_file, 'r', encoding='utf-8') as file: - self.label_list = file.read().split("\n")[:-1] - self.classification = self.classify - self._set_config() - - def get_expected_image_width(self): - return 224 - - def get_expected_image_height(self): - return 224 - - def get_pretrained_images_mean(self): - im_mean = np.array([0.485, 0.456, 0.406]).reshape(1, 3) - return im_mean - - def get_pretrained_images_std(self): - im_std = np.array([0.229, 0.224, 0.225]).reshape(1, 3) - return im_std - - def _set_config(self): - """ - predictor config setting +import math +import collections +import re +import copy + +import paddle +from paddle import ParamAttr +import paddle.nn as nn +import paddle.nn.functional as F +from paddle.nn import Conv2d, BatchNorm, Linear, Dropout +from paddle.nn import AdaptiveAvgPool2d, MaxPool2d, AvgPool2d +from paddlehub.module.module import moduleinfo +from paddlehub.module.cv_module import ImageClassifierModule + +GlobalParams = collections.namedtuple('GlobalParams', [ + 'batch_norm_momentum', + 'batch_norm_epsilon', + 'dropout_rate', + 'num_classes', + 'width_coefficient', + 'depth_coefficient', + 'depth_divisor', + 'min_depth', + 'drop_connect_rate', +]) + +BlockArgs = collections.namedtuple( + 'BlockArgs', + ['kernel_size', 'num_repeat', 'input_filters', 'output_filters', 'expand_ratio', 'id_skip', 'stride', 'se_ratio']) + +GlobalParams.__new__.__defaults__ = (None, ) * len(GlobalParams._fields) +BlockArgs.__new__.__defaults__ = (None, ) * len(BlockArgs._fields) + + +def efficientnet_params(model_name: str): + """ Map EfficientNet model name to parameter coefficients. """ + params_dict = { + # Coefficients: width,depth,resolution,dropout + 'efficientnet-b6': (1.8, 2.6, 528, 0.5) + } + return params_dict[model_name] + + +def efficientnet(width_coefficient: float = None, + depth_coefficient: float = None, + dropout_rate: float = 0.2, + drop_connect_rate: float = 0.2): + """ Get block arguments according to parameter and coefficients. """ + blocks_args = [ + 'r1_k3_s11_e1_i32_o16_se0.25', + 'r2_k3_s22_e6_i16_o24_se0.25', + 'r2_k5_s22_e6_i24_o40_se0.25', + 'r3_k3_s22_e6_i40_o80_se0.25', + 'r3_k5_s11_e6_i80_o112_se0.25', + 'r4_k5_s22_e6_i112_o192_se0.25', + 'r1_k3_s11_e6_i192_o320_se0.25', + ] + blocks_args = BlockDecoder.decode(blocks_args) + + global_params = GlobalParams(batch_norm_momentum=0.99, + batch_norm_epsilon=1e-3, + dropout_rate=dropout_rate, + drop_connect_rate=drop_connect_rate, + num_classes=1000, + width_coefficient=width_coefficient, + depth_coefficient=depth_coefficient, + depth_divisor=8, + min_depth=None) + + return blocks_args, global_params + + +def get_model_params(model_name: str, override_params: dict): + """ Get the block args and global params for a given model """ + if model_name.startswith('efficientnet'): + w, d, _, p = efficientnet_params(model_name) + blocks_args, global_params = efficientnet(width_coefficient=w, depth_coefficient=d, dropout_rate=p) + else: + raise NotImplementedError('model name is not pre-defined: %s' % model_name) + if override_params: + global_params = global_params._replace(**override_params) + return blocks_args, global_params + + +def round_filters(filters: int, global_params: dict): + """ Calculate and round number of filters based on depth multiplier. """ + multiplier = global_params.width_coefficient + if not multiplier: + return filters + divisor = global_params.depth_divisor + min_depth = global_params.min_depth + filters *= multiplier + min_depth = min_depth or divisor + new_filters = max(min_depth, int(filters + divisor / 2) // divisor * divisor) + if new_filters < 0.9 * filters: # prevent rounding by more than 10% + new_filters += divisor + return int(new_filters) + + +def round_repeats(repeats: int, global_params: dict): + """ Round number of filters based on depth multiplier. """ + multiplier = global_params.depth_coefficient + if not multiplier: + return repeats + return int(math.ceil(multiplier * repeats)) + + +class BlockDecoder(object): + """ + Block Decoder, straight from the official TensorFlow repository. + """ + @staticmethod + def _decode_block_string(block_string: str): + """ Gets a block through a string notation of arguments. """ + assert isinstance(block_string, str) + + ops = block_string.split('_') + options = {} + for op in ops: + splits = re.split(r'(\d.*)', op) + if len(splits) >= 2: + key, value = splits[:2] + options[key] = value + + # Check stride + cond_1 = ('s' in options and len(options['s']) == 1) + cond_2 = ((len(options['s']) == 2) and (options['s'][0] == options['s'][1])) + assert (cond_1 or cond_2) + + return BlockArgs(kernel_size=int(options['k']), + num_repeat=int(options['r']), + input_filters=int(options['i']), + output_filters=int(options['o']), + expand_ratio=int(options['e']), + id_skip=('noskip' not in block_string), + se_ratio=float(options['se']) if 'se' in options else None, + stride=[int(options['s'][0])]) + + @staticmethod + def _encode_block_string(block): + """Encodes a block to a string.""" + args = [ + 'r%d' % block.num_repeat, + 'k%d' % block.kernel_size, + 's%d%d' % (block.strides[0], block.strides[1]), + 'e%s' % block.expand_ratio, + 'i%d' % block.input_filters, + 'o%d' % block.output_filters + ] + if 0 < block.se_ratio <= 1: + args.append('se%s' % block.se_ratio) + if block.id_skip is False: + args.append('noskip') + return '_'.join(args) + + @staticmethod + def decode(string_list: list): """ - cpu_config = AnalysisConfig(self.default_pretrained_model_path) - cpu_config.disable_glog_info() - cpu_config.disable_gpu() - self.cpu_predictor = create_paddle_predictor(cpu_config) - - try: - _places = os.environ["CUDA_VISIBLE_DEVICES"] - int(_places[0]) - use_gpu = True - except: - use_gpu = False - if use_gpu: - gpu_config = AnalysisConfig(self.default_pretrained_model_path) - gpu_config.disable_glog_info() - gpu_config.enable_use_gpu( - memory_pool_init_size_mb=1000, device_id=0) - self.gpu_predictor = create_paddle_predictor(gpu_config) - - def context(self, - trainable=True, - pretrained=True, - override_params=None, - phase='train'): - """context for transfer learning. - - Args: - trainable (bool): Set parameters in program to be trainable. - pretrained (bool) : Whether to load pretrained model. - - Returns: - inputs (dict): key is 'image', corresponding vaule is image tensor. - outputs (dict): key is : - 'classification', corresponding value is the result of classification. - 'feature_map', corresponding value is the result of the layer before the fully connected layer. - context_prog (fluid.Program): program for transfer learning. - """ - if phase in ["dev", "test", "predict", "eval"]: - is_test = False - elif phase in ["train"]: - is_test = True - else: - raise ValueError( - "Phase %s is error, which must be one of train, dev, test, eval and predict." - % phase) - - context_prog = fluid.Program() - startup_prog = fluid.Program() - with fluid.program_guard(context_prog, startup_prog): - with fluid.unique_name.guard(): - image = fluid.layers.data( - name="image", shape=[3, 224, 224], dtype="float32") - efficientnet_b6 = EfficientNetB6( - override_params=override_params) - output, feature_map = efficientnet_b6.net( - input=image, - class_dim=len(self.label_list), - is_test=is_test) - - name_prefix = '@HUB_{}@'.format(self.name) - inputs = {'image': name_prefix + image.name} - outputs = { - 'classification': name_prefix + output.name, - 'feature_map': name_prefix + feature_map.name - } - add_vars_prefix(context_prog, name_prefix) - add_vars_prefix(startup_prog, name_prefix) - global_vars = context_prog.global_block().vars - inputs = { - key: global_vars[value] - for key, value in inputs.items() - } - outputs = { - key: global_vars[value] - for key, value in outputs.items() - } - - place = fluid.CPUPlace() - exe = fluid.Executor(place) - # pretrained - if pretrained: - - def _if_exist(var): - b = os.path.exists( - os.path.join(self.default_pretrained_model_path, - var.name)) - return b - - fluid.io.load_vars( - exe, - self.default_pretrained_model_path, - context_prog, - predicate=_if_exist) - else: - exe.run(startup_prog) - # trainable - for param in context_prog.global_block().iter_parameters(): - param.trainable = trainable - return inputs, outputs, context_prog - - def classify(self, - images=None, - paths=None, - batch_size=1, - use_gpu=False, - top_k=1): - """ - API for image classification. - - Args: - images (list[numpy.ndarray]): data of images, shape of each is [H, W, C], color space must be BGR. - paths (list[str]): The paths of images. - batch_size (int): batch size. - use_gpu (bool): Whether to use gpu. - top_k (int): Return top k results. + Decode a list of string notations to specify blocks in the network. - Returns: - res (list[dict]): The classfication results. + string_list: list of strings, each string is a notation of block + return + list of BlockArgs namedtuples of block args """ - if use_gpu: - try: - _places = os.environ["CUDA_VISIBLE_DEVICES"] - int(_places[0]) - except: - raise RuntimeError( - "Environment Variable CUDA_VISIBLE_DEVICES is not set correctly. If you wanna use gpu, please set CUDA_VISIBLE_DEVICES as cuda_device_id." - ) - - all_data = list() - for yield_data in reader(images, paths): - all_data.append(yield_data) - - total_num = len(all_data) - loop_num = int(np.ceil(total_num / batch_size)) - - res = list() - for iter_id in range(loop_num): - batch_data = list() - handle_id = iter_id * batch_size - for image_id in range(batch_size): - try: - batch_data.append(all_data[handle_id + image_id]) - except: - pass - # feed batch image - batch_image = np.array([data['image'] for data in batch_data]) - batch_image = PaddleTensor(batch_image.copy()) - predictor_output = self.gpu_predictor.run([ - batch_image - ]) if use_gpu else self.cpu_predictor.run([batch_image]) - out = postprocess( - data_out=predictor_output[0].as_ndarray(), - label_list=self.label_list, - top_k=top_k) - res += out - return res - - def save_inference_model(self, - dirname, - model_filename=None, - params_filename=None, - combined=True): - if combined: - model_filename = "__model__" if not model_filename else model_filename - params_filename = "__params__" if not params_filename else params_filename - place = fluid.CPUPlace() - exe = fluid.Executor(place) - - program, feeded_var_names, target_vars = fluid.io.load_inference_model( - dirname=self.default_pretrained_model_path, executor=exe) - - fluid.io.save_inference_model( - dirname=dirname, - main_program=program, - executor=exe, - feeded_var_names=feeded_var_names, - target_vars=target_vars, - model_filename=model_filename, - params_filename=params_filename) - - @serving - def serving_method(self, images, **kwargs): + assert isinstance(string_list, list) + blocks_args = [] + for block_string in string_list: + blocks_args.append(BlockDecoder._decode_block_string(block_string)) + return blocks_args + + @staticmethod + def encode(blocks_args: list): """ - Run as a service. - """ - images_decode = [base64_to_cv2(image) for image in images] - results = self.classify(images=images_decode, **kwargs) - return results + Encodes a list of BlockArgs to a list of strings. - @runnable - def run_cmd(self, argvs): - """ - Run as a command. + :param blocks_args: a list of BlockArgs namedtuples of block args + :return: a list of strings, each string is a notation of block """ - self.parser = argparse.ArgumentParser( - description="Run the {} module.".format(self.name), - prog='hub run {}'.format(self.name), - usage='%(prog)s', - add_help=True) - self.arg_input_group = self.parser.add_argument_group( - title="Input options", description="Input data. Required") - self.arg_config_group = self.parser.add_argument_group( - title="Config options", - description= - "Run configuration for controlling module behavior, not required.") - self.add_module_config_arg() - self.add_module_input_arg() - args = self.parser.parse_args(argvs) - results = self.classify( - paths=[args.input_path], - batch_size=args.batch_size, - use_gpu=args.use_gpu) - return results - - def add_module_config_arg(self): - """ - Add the command config options. - """ - self.arg_config_group.add_argument( - '--use_gpu', - type=ast.literal_eval, - default=False, - help="whether use GPU or not.") - self.arg_config_group.add_argument( - '--batch_size', - type=ast.literal_eval, - default=1, - help="batch size.") - self.arg_config_group.add_argument( - '--top_k', - type=ast.literal_eval, - default=1, - help="Return top k results.") - - def add_module_input_arg(self): - """ - Add the command input options. - """ - self.arg_input_group.add_argument( - '--input_path', type=str, help="path to image.") - - -if __name__ == '__main__': - b6 = EfficientNetB6ImageNet() - b6.context() - import cv2 - test_image = [cv2.imread('dog.jpeg')] - res = b6.classification(images=test_image) - print(res) - res = b6.classification(paths=['dog.jpeg']) - print(res) - res = b6.classification(images=test_image) - print(res) - res = b6.classify(images=test_image) - print(res) + block_strings = [] + for block in blocks_args: + block_strings.append(BlockDecoder._encode_block_string(block)) + return block_strings + + +def initial_type(name: str, use_bias: bool = False): + param_attr = ParamAttr(name=name + "_weights") + if use_bias: + bias_attr = ParamAttr(name=name + "_offset") + else: + bias_attr = False + return param_attr, bias_attr + + +def init_batch_norm_layer(name: str = "batch_norm"): + param_attr = ParamAttr(name=name + "_scale") + bias_attr = ParamAttr(name=name + "_offset") + return param_attr, bias_attr + + +def init_fc_layer(name: str = "fc"): + param_attr = ParamAttr(name=name + "_weights") + bias_attr = ParamAttr(name=name + "_offset") + return param_attr, bias_attr + + +def cal_padding(img_size: int, stride: int, filter_size: int, dilation: int = 1): + """Calculate padding size.""" + if img_size % stride == 0: + out_size = max(filter_size - stride, 0) + else: + out_size = max(filter_size - (img_size % stride), 0) + return out_size // 2, out_size - out_size // 2 + + +inp_shape = {"b6": [528, 264, 264, 132, 66, 33, 33, 17]} + + +def _drop_connect(inputs: paddle.Tensor, prob: float, is_test: bool): + """Drop input connection""" + if is_test: + return inputs + keep_prob = 1.0 - prob + inputs_shape = paddle.shape(inputs) + random_tensor = keep_prob + paddle.rand(shape=[inputs_shape[0], 1, 1, 1]) + binary_tensor = paddle.floor(random_tensor) + output = inputs / keep_prob * binary_tensor + return output + + +class Conv2ds(nn.Layer): + """Basic conv layer""" + def __init__(self, + input_channels: int, + output_channels: int, + filter_size: int, + stride: int = 1, + padding: int = 0, + groups: int = None, + name: str = "conv2d", + act: str = None, + use_bias: bool = False, + padding_type: str = None, + model_name: str = None, + cur_stage: str = None): + super(Conv2ds, self).__init__() + assert act in [None, "swish", "sigmoid"] + self.act = act + + param_attr, bias_attr = initial_type(name=name, use_bias=use_bias) + + def get_padding(filter_size, stride=1, dilation=1): + padding = ((stride - 1) + dilation * (filter_size - 1)) // 2 + return padding + + inps = 1 if model_name == None and cur_stage == None else inp_shape[model_name][cur_stage] + self.need_crop = False + if padding_type == "SAME": + top_padding, bottom_padding = cal_padding(inps, stride, filter_size) + left_padding, right_padding = cal_padding(inps, stride, filter_size) + height_padding = bottom_padding + width_padding = right_padding + if top_padding != bottom_padding or left_padding != right_padding: + height_padding = top_padding + stride + width_padding = left_padding + stride + self.need_crop = True + padding = [height_padding, width_padding] + elif padding_type == "VALID": + height_padding = 0 + width_padding = 0 + padding = [height_padding, width_padding] + elif padding_type == "DYNAMIC": + padding = get_padding(filter_size, stride) + else: + padding = padding_type + + groups = 1 if groups is None else groups + self._conv = Conv2d(input_channels, + output_channels, + filter_size, + groups=groups, + stride=stride, + padding=padding, + weight_attr=param_attr, + bias_attr=bias_attr) + + def forward(self, inputs: paddle.Tensor): + x = self._conv(inputs) + if self.act == "swish": + x = F.swish(x) + elif self.act == "sigmoid": + x = F.sigmoid(x) + + if self.need_crop: + x = x[:, :, 1:, 1:] + return x + + +class ConvBNLayer(nn.Layer): + """Basic conv bn layer.""" + def __init__(self, + input_channels: int, + filter_size: int, + output_channels: int, + stride: int = 1, + num_groups: int = 1, + padding_type: str = "SAME", + conv_act: str = None, + bn_act: str = "swish", + use_bn: bool = True, + use_bias: bool = False, + name: str = None, + conv_name: str = None, + bn_name: str = None, + model_name: str = None, + cur_stage: str = None): + super(ConvBNLayer, self).__init__() + + self._conv = Conv2ds(input_channels=input_channels, + output_channels=output_channels, + filter_size=filter_size, + stride=stride, + groups=num_groups, + act=conv_act, + padding_type=padding_type, + name=conv_name, + use_bias=use_bias, + model_name=model_name, + cur_stage=cur_stage) + self.use_bn = use_bn + if use_bn is True: + bn_name = name + bn_name + param_attr, bias_attr = init_batch_norm_layer(bn_name) + + self._bn = BatchNorm(num_channels=output_channels, + act=bn_act, + momentum=0.99, + epsilon=0.001, + moving_mean_name=bn_name + "_mean", + moving_variance_name=bn_name + "_variance", + param_attr=param_attr, + bias_attr=bias_attr) + + def forward(self, inputs: paddle.Tensor): + if self.use_bn: + x = self._conv(inputs) + x = self._bn(x) + return x + else: + return self._conv(inputs) + + +class ExpandConvNorm(nn.Layer): + """Expand conv norm layer.""" + def __init__(self, + input_channels: int, + block_args: dict, + padding_type: str, + name: str = None, + model_name: str = None, + cur_stage: str = None): + super(ExpandConvNorm, self).__init__() + + self.oup = block_args.input_filters * block_args.expand_ratio + self.expand_ratio = block_args.expand_ratio + + if self.expand_ratio != 1: + self._conv = ConvBNLayer(input_channels, + 1, + self.oup, + bn_act=None, + padding_type=padding_type, + name=name, + conv_name=name + "_expand_conv", + bn_name="_bn0", + model_name=model_name, + cur_stage=cur_stage) + + def forward(self, inputs: paddle.Tensor): + if self.expand_ratio != 1: + return self._conv(inputs) + else: + return inputs + + +class DepthwiseConvNorm(nn.Layer): + """Depthwise conv norm layer.""" + def __init__(self, + input_channels: int, + block_args: dict, + padding_type: str, + name: str = None, + model_name: str = None, + cur_stage: str = None): + super(DepthwiseConvNorm, self).__init__() + + self.k = block_args.kernel_size + self.s = block_args.stride + if isinstance(self.s, list) or isinstance(self.s, tuple): + self.s = self.s[0] + oup = block_args.input_filters * block_args.expand_ratio + + self._conv = ConvBNLayer(input_channels, + self.k, + oup, + self.s, + num_groups=input_channels, + bn_act=None, + padding_type=padding_type, + name=name, + conv_name=name + "_depthwise_conv", + bn_name="_bn1", + model_name=model_name, + cur_stage=cur_stage) + + def forward(self, inputs: paddle.Tensor): + return self._conv(inputs) + + +class ProjectConvNorm(nn.Layer): + """Projection conv bn layer.""" + def __init__(self, + input_channels: int, + block_args: dict, + padding_type: str, + name: str = None, + model_name: str = None, + cur_stage: str = None): + super(ProjectConvNorm, self).__init__() + + final_oup = block_args.output_filters + + self._conv = ConvBNLayer(input_channels, + 1, + final_oup, + bn_act=None, + padding_type=padding_type, + name=name, + conv_name=name + "_project_conv", + bn_name="_bn2", + model_name=model_name, + cur_stage=cur_stage) + + def forward(self, inputs: paddle.Tensor): + return self._conv(inputs) + + +class SEBlock(nn.Layer): + """Basic Squeeze-and-Excitation block for Efficientnet.""" + def __init__(self, + input_channels: int, + num_squeezed_channels: int, + oup: int, + padding_type: str, + name: str = None, + model_name: str = None, + cur_stage: str = None): + super(SEBlock, self).__init__() + + self._pool = AdaptiveAvgPool2d(1) + self._conv1 = Conv2ds(input_channels, + num_squeezed_channels, + 1, + use_bias=True, + padding_type=padding_type, + act="swish", + name=name + "_se_reduce") + + self._conv2 = Conv2ds(num_squeezed_channels, + oup, + 1, + act="sigmoid", + use_bias=True, + padding_type=padding_type, + name=name + "_se_expand") + + def forward(self, inputs: paddle.Tensor): + x = self._pool(inputs) + x = self._conv1(x) + x = self._conv2(x) + return paddle.multiply(inputs, x) + + +class MbConvBlock(nn.Layer): + """Mobile inverted bottleneck convolution for Efficientnet.""" + def __init__(self, + input_channels: int, + block_args: dict, + padding_type: str, + use_se: bool, + name: str = None, + drop_connect_rate: float = None, + is_test: bool = False, + model_name: str = None, + cur_stage: str = None): + super(MbConvBlock, self).__init__() + + oup = block_args.input_filters * block_args.expand_ratio + self.block_args = block_args + self.has_se = use_se and (block_args.se_ratio is not None) and (0 < block_args.se_ratio <= 1) + self.id_skip = block_args.id_skip + self.expand_ratio = block_args.expand_ratio + self.drop_connect_rate = drop_connect_rate + self.is_test = is_test + + if self.expand_ratio != 1: + self._ecn = ExpandConvNorm(input_channels, + block_args, + padding_type=padding_type, + name=name, + model_name=model_name, + cur_stage=cur_stage) + + self._dcn = DepthwiseConvNorm(input_channels * block_args.expand_ratio, + block_args, + padding_type=padding_type, + name=name, + model_name=model_name, + cur_stage=cur_stage) + + if self.has_se: + num_squeezed_channels = max(1, int(block_args.input_filters * block_args.se_ratio)) + self._se = SEBlock(input_channels * block_args.expand_ratio, + num_squeezed_channels, + oup, + padding_type=padding_type, + name=name, + model_name=model_name, + cur_stage=cur_stage) + + self._pcn = ProjectConvNorm(input_channels * block_args.expand_ratio, + block_args, + padding_type=padding_type, + name=name, + model_name=model_name, + cur_stage=cur_stage) + + def forward(self, inputs: paddle.Tensor): + x = inputs + if self.expand_ratio != 1: + x = self._ecn(x) + x = F.swish(x) + x = self._dcn(x) + x = F.swish(x) + if self.has_se: + x = self._se(x) + x = self._pcn(x) + if self.id_skip and \ + self.block_args.stride == 1 and \ + self.block_args.input_filters == self.block_args.output_filters: + if self.drop_connect_rate: + x = _drop_connect(x, self.drop_connect_rate, self.is_test) + x = paddle.elementwise_add(x, inputs) + return x + + +class ConvStemNorm(nn.Layer): + """Basic conv stem norm block for extracting features.""" + def __init__(self, + input_channels: int, + padding_type: str, + _global_params: dict, + name: str = None, + model_name: str = None, + cur_stage: str = None): + super(ConvStemNorm, self).__init__() + + output_channels = round_filters(32, _global_params) + self._conv = ConvBNLayer(input_channels, + filter_size=3, + output_channels=output_channels, + stride=2, + bn_act=None, + padding_type=padding_type, + name="", + conv_name="_conv_stem", + bn_name="_bn0", + model_name=model_name, + cur_stage=cur_stage) + + def forward(self, inputs: paddle.Tensor): + return self._conv(inputs) + + +class ExtractFeatures(nn.Layer): + """Extract features.""" + def __init__(self, + input_channels: int, + _block_args: dict, + _global_params: dict, + padding_type: str, + use_se: bool, + is_test: bool, + model_name: str = None): + super(ExtractFeatures, self).__init__() + + self._global_params = _global_params + + self._conv_stem = ConvStemNorm(input_channels, + padding_type=padding_type, + _global_params=_global_params, + model_name=model_name, + cur_stage=0) + + self.block_args_copy = copy.deepcopy(_block_args) + idx = 0 + block_size = 0 + for block_arg in self.block_args_copy: + block_arg = block_arg._replace(input_filters=round_filters(block_arg.input_filters, _global_params), + output_filters=round_filters(block_arg.output_filters, _global_params), + num_repeat=round_repeats(block_arg.num_repeat, _global_params)) + block_size += 1 + for _ in range(block_arg.num_repeat - 1): + block_size += 1 + + self.conv_seq = [] + cur_stage = 1 + for block_args in _block_args: + block_args = block_args._replace(input_filters=round_filters(block_args.input_filters, _global_params), + output_filters=round_filters(block_args.output_filters, _global_params), + num_repeat=round_repeats(block_args.num_repeat, _global_params)) + + drop_connect_rate = self._global_params.drop_connect_rate if not is_test else 0 + if drop_connect_rate: + drop_connect_rate *= float(idx) / block_size + + _mc_block = self.add_sublayer( + "_blocks." + str(idx) + ".", + MbConvBlock(block_args.input_filters, + block_args=block_args, + padding_type=padding_type, + use_se=use_se, + name="_blocks." + str(idx) + ".", + drop_connect_rate=drop_connect_rate, + model_name=model_name, + cur_stage=cur_stage)) + self.conv_seq.append(_mc_block) + idx += 1 + if block_args.num_repeat > 1: + block_args = block_args._replace(input_filters=block_args.output_filters, stride=1) + for _ in range(block_args.num_repeat - 1): + drop_connect_rate = self._global_params.drop_connect_rate if not is_test else 0 + if drop_connect_rate: + drop_connect_rate *= float(idx) / block_size + _mc_block = self.add_sublayer( + "block." + str(idx) + ".", + MbConvBlock(block_args.input_filters, + block_args, + padding_type=padding_type, + use_se=use_se, + name="_blocks." + str(idx) + ".", + drop_connect_rate=drop_connect_rate, + model_name=model_name, + cur_stage=cur_stage)) + self.conv_seq.append(_mc_block) + idx += 1 + cur_stage += 1 + + def forward(self, inputs: paddle.Tensor): + x = self._conv_stem(inputs) + x = F.swish(x) + for _mc_block in self.conv_seq: + x = _mc_block(x) + return x + + +@moduleinfo(name="efficientnetb6_imagenet", + type="cv/classification", + author="paddlepaddle", + author_email="", + summary="efficientnetb6_imagenet is a classification model, " + "this module is trained with Imagenet dataset.", + version="1.1.0", + meta=ImageClassifierModule) +class EfficientNet_B6(nn.Layer): + def __init__(self, + is_test: bool = False, + padding_type: str = "SAME", + override_params: dict = None, + use_se: bool = True, + class_dim: int = 1000, + load_checkpoint: str = None): + super(EfficientNet_B6, self).__init__() + + model_name = 'efficientnet-b6' + self.name = "b6" + self._block_args, self._global_params = get_model_params(model_name, override_params) + self.padding_type = padding_type + self.use_se = use_se + self.is_test = is_test + + self._ef = ExtractFeatures(3, + self._block_args, + self._global_params, + self.padding_type, + self.use_se, + self.is_test, + model_name=self.name) + + output_channels = round_filters(1280, self._global_params) + oup = 576 + + self._conv = ConvBNLayer(oup, + 1, + output_channels, + bn_act="swish", + padding_type=self.padding_type, + name="", + conv_name="_conv_head", + bn_name="_bn1", + model_name=self.name, + cur_stage=7) + self._pool = AdaptiveAvgPool2d(1) + + if self._global_params.dropout_rate: + self._drop = Dropout(p=self._global_params.dropout_rate, mode="upscale_in_train") + + param_attr, bias_attr = init_fc_layer("_fc") + self._fc = Linear(output_channels, class_dim, weight_attr=param_attr, bias_attr=bias_attr) + + if load_checkpoint is not None: + model_dict = paddle.load(load_checkpoint)[0] + self.set_dict(model_dict) + print("load custom checkpoint success") + + else: + checkpoint = os.path.join(self.directory, 'efficientnet_b6_imagenet.pdparams') + if not os.path.exists(checkpoint): + os.system( + 'wget https://paddlehub.bj.bcebos.com/dygraph/image_classification/efficientnet_b6_imagenet.pdparams -O ' + + checkpoint) + model_dict = paddle.load(checkpoint)[0] + self.set_dict(model_dict) + print("load pretrained checkpoint success") + + def forward(self, inputs: paddle.Tensor): + x = self._ef(inputs) + x = self._conv(x) + x = self._pool(x) + if self._global_params.dropout_rate: + x = self._drop(x) + x = paddle.squeeze(x, axis=[2, 3]) + x = self._fc(x) + return x diff --git a/hub_module/modules/image/classification/efficientnetb6_imagenet/processor.py b/hub_module/modules/image/classification/efficientnetb6_imagenet/processor.py deleted file mode 100644 index 12ccc9b2..00000000 --- a/hub_module/modules/image/classification/efficientnetb6_imagenet/processor.py +++ /dev/null @@ -1,69 +0,0 @@ -# -*- coding:utf-8 -*- -# copyright (c) 2020 PaddlePaddle Authors. All Rights Reserve. -# -# Licensed under the Apache License, Version 2.0 (the "License"); -# you may not use this file except in compliance with the License. -# You may obtain a copy of the License at -# -# http://www.apache.org/licenses/LICENSE-2.0 -# -# Unless required by applicable law or agreed to in writing, software -# distributed under the License is distributed on an "AS IS" BASIS, -# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. -# See the License for the specific language governing permissions and -# limitations under the License. - -from __future__ import absolute_import -from __future__ import division -from __future__ import print_function - -import base64 -import cv2 -import os - -import numpy as np - - -def base64_to_cv2(b64str): - data = base64.b64decode(b64str.encode('utf8')) - data = np.fromstring(data, np.uint8) - data = cv2.imdecode(data, cv2.IMREAD_COLOR) - return data - - -def softmax(x): - orig_shape = x.shape - if len(x.shape) > 1: - tmp = np.max(x, axis=1) - x -= tmp.reshape((x.shape[0], 1)) - x = np.exp(x) - tmp = np.sum(x, axis=1) - x /= tmp.reshape((x.shape[0], 1)) - else: - tmp = np.max(x) - x -= tmp - x = np.exp(x) - tmp = np.sum(x) - x /= tmp - return x - - -def postprocess(data_out, label_list, top_k): - """ - Postprocess output of network, one image at a time. - - Args: - data_out (numpy.ndarray): output data of network. - label_list (list): list of label. - top_k (int): Return top k results. - """ - output = [] - for result in data_out: - result_i = softmax(result) - output_i = {} - indexs = np.argsort(result_i)[::-1][0:top_k] - for index in indexs: - label = label_list[index].split(',')[0] - output_i[label] = float(result_i[index]) - output.append(output_i) - return output diff --git a/hub_module/modules/image/classification/efficientnetb7_imagenet/README.md b/hub_module/modules/image/classification/efficientnetb7_imagenet/README.md deleted file mode 100644 index 75e9d397..00000000 --- a/hub_module/modules/image/classification/efficientnetb7_imagenet/README.md +++ /dev/null @@ -1,149 +0,0 @@ -## 命令行预测 - -``` -hub run efficientnetb7_imagenet --input_path "/PATH/TO/IMAGE" -``` - -## API - -```python -def get_expected_image_width() -``` - -返回预处理的图片宽度，也就是224。 - -```python -def get_expected_image_height() -``` - -返回预处理的图片高度，也就是224。 - -```python -def get_pretrained_images_mean() -``` - -返回预处理的图片均值，也就是 \[0.485, 0.456, 0.406\]。 - -```python -def get_pretrained_images_std() -``` - -返回预处理的图片标准差，也就是 \[0.229, 0.224, 0.225\]。 - - -```python -def context(trainable=True, pretrained=True) -``` - -**参数** - -* trainable (bool): 计算图的参数是否为可训练的； -* pretrained (bool): 是否加载默认的预训练模型。 - -**返回** - -* inputs (dict): 计算图的输入，key 为 'image', value 为图片的张量； -* outputs (dict): 计算图的输出，key 为 'classification' 和 'feature_map'，其相应的值为： - * classification (paddle.fluid.framework.Variable): 分类结果，也就是全连接层的输出； - * feature\_map (paddle.fluid.framework.Variable): 特征匹配，全连接层前面的那个张量。 -* context\_prog(fluid.Program): 计算图，用于迁移学习。 - -```python -def classify(images=None, - paths=None, - batch_size=1, - use_gpu=False, - top_k=1): -``` - -**参数** - -* images (list\[numpy.ndarray\]): 图片数据，每一个图片数据的shape 均为 \[H, W, C\]，颜色空间为 BGR； -* paths (list\[str\]): 图片的路径； -* batch\_size (int): batch 的大小； -* use\_gpu (bool): 是否使用 GPU 来预测； -* top\_k (int): 返回预测结果的前 k 个。 - -**返回** - -res (list\[dict\]): 分类结果，列表的每一个元素均为字典，其中 key 为识别动物的类别，value为置信度。 - -```python -def save_inference_model(dirname, - model_filename=None, - params_filename=None, - combined=True) -``` - -将模型保存到指定路径。 - -**参数** - -* dirname: 存在模型的目录名称 -* model\_filename: 模型文件名称，默认为\_\_model\_\_ -* params\_filename: 参数文件名称，默认为\_\_params\_\_(仅当`combined`为True时生效) -* combined: 是否将参数保存到统一的一个文件中 - -## 代码示例 - -```python -import paddlehub as hub -import cv2 - -classifier = hub.Module(name="efficientnetb7_imagenet") - -result = classifier.classify(images=[cv2.imread('/PATH/TO/IMAGE')]) -# or -# result = classifier.classify(paths=['/PATH/TO/IMAGE']) -``` - -## 服务部署 - -PaddleHub Serving可以部署一个在线图像识别服务。 - -## 第一步：启动PaddleHub Serving - -运行启动命令： -```shell -$ hub serving start -m efficientnetb7_imagenet -``` - -这样就完成了一个在线图像识别服务化API的部署，默认端口号为8866。 - -**NOTE:** 如使用GPU预测，则需要在启动服务之前，请设置CUDA\_VISIBLE\_DEVICES环境变量，否则不用设置。 - -## 第二步：发送预测请求 - -配置好服务端，以下数行代码即可实现发送预测请求，获取预测结果 - -```python -import requests -import json -import cv2 -import base64 - - -def cv2_to_base64(image): - data = cv2.imencode('.jpg', image)[1] - return base64.b64encode(data.tostring()).decode('utf8') - - -# 发送HTTP请求 -data = {'images':[cv2_to_base64(cv2.imread("/PATH/TO/IMAGE"))]} -headers = {"Content-type": "application/json"} -url = "http://127.0.0.1:8866/predict/efficientnetb7_imagenet" -r = requests.post(url=url, headers=headers, data=json.dumps(data)) - -# 打印预测结果 -print(r.json()["results"]) -``` - -### 查看代码 - -https://github.com/PaddlePaddle/PaddleClas - -### 依赖 - -paddlepaddle >= 1.6.2 - -paddlehub >= 1.6.0 diff --git a/hub_module/modules/image/classification/efficientnetb7_imagenet/__init__.py b/hub_module/modules/image/classification/efficientnetb7_imagenet/__init__.py deleted file mode 100644 index 19a990da..00000000 --- a/hub_module/modules/image/classification/efficientnetb7_imagenet/__init__.py +++ /dev/null @@ -1 +0,0 @@ -# -*- coding:utf-8 -*- diff --git a/hub_module/modules/image/classification/efficientnetb7_imagenet/data_feed.py b/hub_module/modules/image/classification/efficientnetb7_imagenet/data_feed.py deleted file mode 100644 index 11181458..00000000 --- a/hub_module/modules/image/classification/efficientnetb7_imagenet/data_feed.py +++ /dev/null @@ -1,98 +0,0 @@ -# -*- coding:utf-8 -*- -# copyright (c) 2020 PaddlePaddle Authors. All Rights Reserve. -# -# Licensed under the Apache License, Version 2.0 (the "License"); -# you may not use this file except in compliance with the License. -# You may obtain a copy of the License at -# -# http://www.apache.org/licenses/LICENSE-2.0 -# -# Unless required by applicable law or agreed to in writing, software -# distributed under the License is distributed on an "AS IS" BASIS, -# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. -# See the License for the specific language governing permissions and -# limitations under the License. - -import os -import time -from collections import OrderedDict - -import cv2 -import numpy as np -from PIL import Image - -__all__ = ['reader'] - -DATA_DIM = 224 -img_mean = np.array([0.485, 0.456, 0.406]).reshape((3, 1, 1)) -img_std = np.array([0.229, 0.224, 0.225]).reshape((3, 1, 1)) - - -def resize_short(img, target_size): - percent = float(target_size) / min(img.size[0], img.size[1]) - resized_width = int(round(img.size[0] * percent)) - resized_height = int(round(img.size[1] * percent)) - img = img.resize((resized_width, resized_height), Image.LANCZOS) - return img - - -def crop_image(img, target_size, center): - width, height = img.size - size = target_size - if center == True: - w_start = (width - size) / 2 - h_start = (height - size) / 2 - else: - w_start = np.random.randint(0, width - size + 1) - h_start = np.random.randint(0, height - size + 1) - w_end = w_start + size - h_end = h_start + size - img = img.crop((w_start, h_start, w_end, h_end)) - return img - - -def process_image(img): - img = resize_short(img, target_size=256) - img = crop_image(img, target_size=DATA_DIM, center=True) - if img.mode != 'RGB': - img = img.convert('RGB') - img = np.array(img).astype('float32').transpose((2, 0, 1)) / 255 - img -= img_mean - img /= img_std - return img - - -def reader(images=None, paths=None): - """ - Preprocess to yield image. - - Args: - images (list[numpy.ndarray]): images data, shape of each is [H, W, C]. - paths (list[str]): paths to images. - - Yield: - each (collections.OrderedDict): info of original image, preprocessed image. - """ - component = list() - if paths: - for im_path in paths: - each = OrderedDict() - assert os.path.isfile( - im_path), "The {} isn't a valid file path.".format(im_path) - each['org_im_path'] = im_path - each['org_im'] = Image.open(im_path) - each['org_im_width'], each['org_im_height'] = each['org_im'].size - component.append(each) - if images is not None: - assert type(images), "images is a list." - for im in images: - each = OrderedDict() - each['org_im'] = Image.fromarray(im[:, :, ::-1]) - each['org_im_path'] = 'ndarray_time={}'.format( - round(time.time(), 6) * 1e6) - each['org_im_width'], each['org_im_height'] = each['org_im'].size - component.append(each) - - for element in component: - element['image'] = process_image(element['org_im']) - yield element diff --git a/hub_module/modules/image/classification/efficientnetb7_imagenet/efficientnet.py b/hub_module/modules/image/classification/efficientnetb7_imagenet/efficientnet.py deleted file mode 100644 index 7024c8ad..00000000 --- a/hub_module/modules/image/classification/efficientnetb7_imagenet/efficientnet.py +++ /dev/null @@ -1,623 +0,0 @@ -# -*- coding:utf-8 -*- -# copyright (c) 2020 PaddlePaddle Authors. All Rights Reserve. -# -# Licensed under the Apache License, Version 2.0 (the "License"); -# you may not use this file except in compliance with the License. -# You may obtain a copy of the License at -# -# http://www.apache.org/licenses/LICENSE-2.0 -# -# Unless required by applicable law or agreed to in writing, software -# distributed under the License is distributed on an "AS IS" BASIS, -# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. -# See the License for the specific language governing permissions and -# limitations under the License. - -from __future__ import absolute_import -from __future__ import division -from __future__ import print_function - -import collections -import re -import math -import copy - -import paddle.fluid as fluid -from efficientnetb7_imagenet.layers import conv2d, init_batch_norm_layer, init_fc_layer - -__all__ = [ - 'EfficientNet', 'EfficientNetB0_small', 'EfficientNetB0', 'EfficientNetB1', - 'EfficientNetB2', 'EfficientNetB3', 'EfficientNetB4', 'EfficientNetB5', - 'EfficientNetB6', 'EfficientNetB7' -] - -GlobalParams = collections.namedtuple('GlobalParams', [ - 'batch_norm_momentum', - 'batch_norm_epsilon', - 'dropout_rate', - 'num_classes', - 'width_coefficient', - 'depth_coefficient', - 'depth_divisor', - 'min_depth', - 'drop_connect_rate', -]) - -BlockArgs = collections.namedtuple('BlockArgs', [ - 'kernel_size', 'num_repeat', 'input_filters', 'output_filters', - 'expand_ratio', 'id_skip', 'stride', 'se_ratio' -]) - -GlobalParams.__new__.__defaults__ = (None, ) * len(GlobalParams._fields) -BlockArgs.__new__.__defaults__ = (None, ) * len(BlockArgs._fields) - - -def efficientnet_params(model_name): - """ Map EfficientNet model name to parameter coefficients. """ - params_dict = { - # Coefficients: width,depth,resolution,dropout - 'efficientnet-b0': (1.0, 1.0, 224, 0.2), - 'efficientnet-b1': (1.0, 1.1, 240, 0.2), - 'efficientnet-b2': (1.1, 1.2, 260, 0.3), - 'efficientnet-b3': (1.2, 1.4, 300, 0.3), - 'efficientnet-b4': (1.4, 1.8, 380, 0.4), - 'efficientnet-b5': (1.6, 2.2, 456, 0.4), - 'efficientnet-b6': (1.8, 2.6, 528, 0.5), - 'efficientnet-b7': (2.0, 3.1, 600, 0.5), - } - return params_dict[model_name] - - -def efficientnet(width_coefficient=None, - depth_coefficient=None, - dropout_rate=0.2, - drop_connect_rate=0.2): - """ Get block arguments according to parameter and coefficients. """ - blocks_args = [ - 'r1_k3_s11_e1_i32_o16_se0.25', - 'r2_k3_s22_e6_i16_o24_se0.25', - 'r2_k5_s22_e6_i24_o40_se0.25', - 'r3_k3_s22_e6_i40_o80_se0.25', - 'r3_k5_s11_e6_i80_o112_se0.25', - 'r4_k5_s22_e6_i112_o192_se0.25', - 'r1_k3_s11_e6_i192_o320_se0.25', - ] - blocks_args = BlockDecoder.decode(blocks_args) - - global_params = GlobalParams( - batch_norm_momentum=0.99, - batch_norm_epsilon=1e-3, - dropout_rate=dropout_rate, - drop_connect_rate=drop_connect_rate, - num_classes=1000, - width_coefficient=width_coefficient, - depth_coefficient=depth_coefficient, - depth_divisor=8, - min_depth=None) - - return blocks_args, global_params - - -def get_model_params(model_name, override_params): - """ Get the block args and global params for a given model """ - if model_name.startswith('efficientnet'): - w, d, _, p = efficientnet_params(model_name) - blocks_args, global_params = efficientnet( - width_coefficient=w, depth_coefficient=d, dropout_rate=p) - else: - raise NotImplementedError( - 'model name is not pre-defined: %s' % model_name) - if override_params: - global_params = global_params._replace(**override_params) - return blocks_args, global_params - - -def round_filters(filters, global_params): - """ Calculate and round number of filters based on depth multiplier. """ - multiplier = global_params.width_coefficient - if not multiplier: - return filters - divisor = global_params.depth_divisor - min_depth = global_params.min_depth - filters *= multiplier - min_depth = min_depth or divisor - new_filters = max(min_depth, - int(filters + divisor / 2) // divisor * divisor) - if new_filters < 0.9 * filters: # prevent rounding by more than 10% - new_filters += divisor - return int(new_filters) - - -def round_repeats(repeats, global_params): - """ Round number of filters based on depth multiplier. """ - multiplier = global_params.depth_coefficient - if not multiplier: - return repeats - return int(math.ceil(multiplier * repeats)) - - -class EfficientNet(): - def __init__(self, - name='b0', - padding_type='SAME', - override_params=None, - is_test=False, - use_se=True): - valid_names = ['b' + str(i) for i in range(8)] - assert name in valid_names, 'efficient name should be in b0~b7' - model_name = 'efficientnet-' + name - self._blocks_args, self._global_params = get_model_params( - model_name, override_params) - self._bn_mom = self._global_params.batch_norm_momentum - self._bn_eps = self._global_params.batch_norm_epsilon - self.padding_type = padding_type - self.use_se = use_se - - def net(self, input, class_dim=1000, is_test=False): - - conv = self.extract_features(input, is_test=is_test) - - out_channels = round_filters(1280, self._global_params) - conv = self.conv_bn_layer( - conv, - num_filters=out_channels, - filter_size=1, - bn_act='swish', - bn_mom=self._bn_mom, - bn_eps=self._bn_eps, - padding_type=self.padding_type, - name='', - conv_name='_conv_head', - bn_name='_bn1') - - pool = fluid.layers.pool2d( - input=conv, pool_type='avg', global_pooling=True, use_cudnn=False) - - if not is_test and self._global_params.dropout_rate: - pool = fluid.layers.dropout( - pool, - self._global_params.dropout_rate, - dropout_implementation='upscale_in_train') - - param_attr, bias_attr = init_fc_layer(class_dim, '_fc') - out = fluid.layers.fc( - pool, - class_dim, - name='_fc', - param_attr=param_attr, - bias_attr=bias_attr) - return out, pool - - def _drop_connect(self, inputs, prob, is_test): - if is_test: - return inputs - keep_prob = 1.0 - prob - random_tensor = keep_prob + fluid.layers.uniform_random_batch_size_like( - inputs, [-1, 1, 1, 1], min=0., max=1.) - binary_tensor = fluid.layers.floor(random_tensor) - output = inputs / keep_prob * binary_tensor - return output - - def _expand_conv_norm(self, inputs, block_args, is_test, name=None): - # Expansion phase - oup = block_args.input_filters * block_args.expand_ratio # number of output channels - - if block_args.expand_ratio != 1: - conv = self.conv_bn_layer( - inputs, - num_filters=oup, - filter_size=1, - bn_act=None, - bn_mom=self._bn_mom, - bn_eps=self._bn_eps, - padding_type=self.padding_type, - name=name, - conv_name=name + '_expand_conv', - bn_name='_bn0') - - return conv - - def _depthwise_conv_norm(self, inputs, block_args, is_test, name=None): - k = block_args.kernel_size - s = block_args.stride - if isinstance(s, list) or isinstance(s, tuple): - s = s[0] - oup = block_args.input_filters * block_args.expand_ratio # number of output channels - - conv = self.conv_bn_layer( - inputs, - num_filters=oup, - filter_size=k, - stride=s, - num_groups=oup, - bn_act=None, - padding_type=self.padding_type, - bn_mom=self._bn_mom, - bn_eps=self._bn_eps, - name=name, - use_cudnn=False, - conv_name=name + '_depthwise_conv', - bn_name='_bn1') - - return conv - - def _project_conv_norm(self, inputs, block_args, is_test, name=None): - final_oup = block_args.output_filters - conv = self.conv_bn_layer( - inputs, - num_filters=final_oup, - filter_size=1, - bn_act=None, - padding_type=self.padding_type, - bn_mom=self._bn_mom, - bn_eps=self._bn_eps, - name=name, - conv_name=name + '_project_conv', - bn_name='_bn2') - return conv - - def conv_bn_layer(self, - input, - filter_size, - num_filters, - stride=1, - num_groups=1, - padding_type="SAME", - conv_act=None, - bn_act='swish', - use_cudnn=True, - use_bn=True, - bn_mom=0.9, - bn_eps=1e-05, - use_bias=False, - name=None, - conv_name=None, - bn_name=None): - conv = conv2d( - input=input, - num_filters=num_filters, - filter_size=filter_size, - stride=stride, - groups=num_groups, - act=conv_act, - padding_type=padding_type, - use_cudnn=use_cudnn, - name=conv_name, - use_bias=use_bias) - - if use_bn == False: - return conv - else: - bn_name = name + bn_name - param_attr, bias_attr = init_batch_norm_layer(bn_name) - return fluid.layers.batch_norm( - input=conv, - act=bn_act, - momentum=bn_mom, - epsilon=bn_eps, - name=bn_name, - moving_mean_name=bn_name + '_mean', - moving_variance_name=bn_name + '_variance', - param_attr=param_attr, - bias_attr=bias_attr) - - def _conv_stem_norm(self, inputs, is_test): - out_channels = round_filters(32, self._global_params) - bn = self.conv_bn_layer( - inputs, - num_filters=out_channels, - filter_size=3, - stride=2, - bn_act=None, - bn_mom=self._bn_mom, - padding_type=self.padding_type, - bn_eps=self._bn_eps, - name='', - conv_name='_conv_stem', - bn_name='_bn0') - - return bn - - def mb_conv_block(self, - inputs, - block_args, - is_test=False, - drop_connect_rate=None, - name=None): - # Expansion and Depthwise Convolution - oup = block_args.input_filters * block_args.expand_ratio # number of output channels - has_se = self.use_se and (block_args.se_ratio is - not None) and (0 < block_args.se_ratio <= 1) - id_skip = block_args.id_skip # skip connection and drop connect - conv = inputs - if block_args.expand_ratio != 1: - conv = fluid.layers.swish( - self._expand_conv_norm(conv, block_args, is_test, name)) - - conv = fluid.layers.swish( - self._depthwise_conv_norm(conv, block_args, is_test, name)) - - # Squeeze and Excitation - if has_se: - num_squeezed_channels = max( - 1, int(block_args.input_filters * block_args.se_ratio)) - conv = self.se_block(conv, num_squeezed_channels, oup, name) - - conv = self._project_conv_norm(conv, block_args, is_test, name) - - # Skip connection and drop connect - input_filters, output_filters = block_args.input_filters, block_args.output_filters - if id_skip and block_args.stride == 1 and input_filters == output_filters: - if drop_connect_rate: - conv = self._drop_connect(conv, drop_connect_rate, is_test) - conv = fluid.layers.elementwise_add(conv, inputs) - - return conv - - def se_block(self, inputs, num_squeezed_channels, oup, name): - x_squeezed = fluid.layers.pool2d( - input=inputs, pool_type='avg', global_pooling=True, use_cudnn=False) - x_squeezed = conv2d( - x_squeezed, - num_filters=num_squeezed_channels, - filter_size=1, - use_bias=True, - padding_type=self.padding_type, - act='swish', - name=name + '_se_reduce') - x_squeezed = conv2d( - x_squeezed, - num_filters=oup, - filter_size=1, - use_bias=True, - padding_type=self.padding_type, - name=name + '_se_expand') - se_out = inputs * fluid.layers.sigmoid(x_squeezed) - return se_out - - def extract_features(self, inputs, is_test): - """ Returns output of the final convolution layer """ - - conv = fluid.layers.swish(self._conv_stem_norm(inputs, is_test=is_test)) - - block_args_copy = copy.deepcopy(self._blocks_args) - idx = 0 - block_size = 0 - for block_arg in block_args_copy: - block_arg = block_arg._replace( - input_filters=round_filters(block_arg.input_filters, - self._global_params), - output_filters=round_filters(block_arg.output_filters, - self._global_params), - num_repeat=round_repeats(block_arg.num_repeat, - self._global_params)) - block_size += 1 - for _ in range(block_arg.num_repeat - 1): - block_size += 1 - - for block_args in self._blocks_args: - - # Update block input and output filters based on depth multiplier. - block_args = block_args._replace( - input_filters=round_filters(block_args.input_filters, - self._global_params), - output_filters=round_filters(block_args.output_filters, - self._global_params), - num_repeat=round_repeats(block_args.num_repeat, - self._global_params)) - - # The first block needs to take care of stride and filter size increase. - drop_connect_rate = self._global_params.drop_connect_rate - if drop_connect_rate: - drop_connect_rate *= float(idx) / block_size - conv = self.mb_conv_block(conv, block_args, is_test, - drop_connect_rate, - '_blocks.' + str(idx) + '.') - - idx += 1 - if block_args.num_repeat > 1: - block_args = block_args._replace( - input_filters=block_args.output_filters, stride=1) - for _ in range(block_args.num_repeat - 1): - drop_connect_rate = self._global_params.drop_connect_rate - if drop_connect_rate: - drop_connect_rate *= float(idx) / block_size - conv = self.mb_conv_block(conv, block_args, is_test, - drop_connect_rate, - '_blocks.' + str(idx) + '.') - idx += 1 - - return conv - - def shortcut(self, input, data_residual): - return fluid.layers.elementwise_add(input, data_residual) - - -class BlockDecoder(object): - """ Block Decoder for readability, straight from the official TensorFlow repository """ - - @staticmethod - def _decode_block_string(block_string): - """ Gets a block through a string notation of arguments. """ - assert isinstance(block_string, str) - - ops = block_string.split('_') - options = {} - for op in ops: - splits = re.split(r'(\d.*)', op) - if len(splits) >= 2: - key, value = splits[:2] - options[key] = value - - # Check stride - assert (('s' in options and len(options['s']) == 1) or - (len(options['s']) == 2 and options['s'][0] == options['s'][1])) - - return BlockArgs( - kernel_size=int(options['k']), - num_repeat=int(options['r']), - input_filters=int(options['i']), - output_filters=int(options['o']), - expand_ratio=int(options['e']), - id_skip=('noskip' not in block_string), - se_ratio=float(options['se']) if 'se' in options else None, - stride=[int(options['s'][0])]) - - @staticmethod - def _encode_block_string(block): - """Encodes a block to a string.""" - args = [ - 'r%d' % block.num_repeat, - 'k%d' % block.kernel_size, - 's%d%d' % (block.strides[0], block.strides[1]), - 'e%s' % block.expand_ratio, - 'i%d' % block.input_filters, - 'o%d' % block.output_filters - ] - if 0 < block.se_ratio <= 1: - args.append('se%s' % block.se_ratio) - if block.id_skip is False: - args.append('noskip') - return '_'.join(args) - - @staticmethod - def decode(string_list): - """ - Decodes a list of string notations to specify blocks inside the network. - :param string_list: a list of strings, each string is a notation of block - :return: a list of BlockArgs namedtuples of block args - """ - assert isinstance(string_list, list) - blocks_args = [] - for block_string in string_list: - blocks_args.append(BlockDecoder._decode_block_string(block_string)) - return blocks_args - - @staticmethod - def encode(blocks_args): - """ - Encodes a list of BlockArgs to a list of strings. - :param blocks_args: a list of BlockArgs namedtuples of block args - :return: a list of strings, each string is a notation of block - """ - block_strings = [] - for block in blocks_args: - block_strings.append(BlockDecoder._encode_block_string(block)) - return block_strings - - -def EfficientNetB0_small(is_test=False, - padding_type='SAME', - override_params=None, - use_se=False): - model = EfficientNet( - name='b0', - is_test=is_test, - padding_type=padding_type, - override_params=override_params, - use_se=use_se) - return model - - -def EfficientNetB0(is_test=False, - padding_type='SAME', - override_params=None, - use_se=True): - model = EfficientNet( - name='b0', - is_test=is_test, - padding_type=padding_type, - override_params=override_params, - use_se=use_se) - return model - - -def EfficientNetB1(is_test=False, - padding_type='SAME', - override_params=None, - use_se=True): - model = EfficientNet( - name='b1', - is_test=is_test, - padding_type=padding_type, - override_params=override_params, - use_se=use_se) - return model - - -def EfficientNetB2(is_test=False, - padding_type='SAME', - override_params=None, - use_se=True): - model = EfficientNet( - name='b2', - is_test=is_test, - padding_type=padding_type, - override_params=override_params, - use_se=use_se) - return model - - -def EfficientNetB3(is_test=False, - padding_type='SAME', - override_params=None, - use_se=True): - model = EfficientNet( - name='b3', - is_test=is_test, - padding_type=padding_type, - override_params=override_params, - use_se=use_se) - return model - - -def EfficientNetB4(is_test=False, - padding_type='SAME', - override_params=None, - use_se=True): - model = EfficientNet( - name='b4', - is_test=is_test, - padding_type=padding_type, - override_params=override_params, - use_se=use_se) - return model - - -def EfficientNetB5(is_test=False, - padding_type='SAME', - override_params=None, - use_se=True): - model = EfficientNet( - name='b5', - is_test=is_test, - padding_type=padding_type, - override_params=override_params, - use_se=use_se) - return model - - -def EfficientNetB6(is_test=False, - padding_type='SAME', - override_params=None, - use_se=True): - model = EfficientNet( - name='b6', - is_test=is_test, - padding_type=padding_type, - override_params=override_params, - use_se=use_se) - return model - - -def EfficientNetB7(is_test=False, - padding_type='SAME', - override_params=None, - use_se=True): - model = EfficientNet( - name='b7', - is_test=is_test, - padding_type=padding_type, - override_params=override_params, - use_se=use_se) - return model diff --git a/hub_module/modules/image/classification/efficientnetb7_imagenet/label_list.txt b/hub_module/modules/image/classification/efficientnetb7_imagenet/label_list.txt deleted file mode 100644 index a509c007..00000000 --- a/hub_module/modules/image/classification/efficientnetb7_imagenet/label_list.txt +++ /dev/null @@ -1,1000 +0,0 @@ -tench, Tinca tinca -goldfish, Carassius auratus -great white shark, white shark, man-eater, man-eating shark, Carcharodon carcharias -tiger shark, Galeocerdo cuvieri -hammerhead, hammerhead shark -electric ray, crampfish, numbfish, torpedo -stingray -cock -hen -ostrich, Struthio camelus -brambling, Fringilla montifringilla -goldfinch, Carduelis carduelis -house finch, linnet, Carpodacus mexicanus -junco, snowbird -indigo bunting, indigo finch, indigo bird, Passerina cyanea -robin, American robin, Turdus migratorius -bulbul -jay -magpie -chickadee -water ouzel, dipper -kite -bald eagle, American eagle, Haliaeetus leucocephalus -vulture -great grey owl, great gray owl, Strix nebulosa -European fire salamander, Salamandra salamandra -common newt, Triturus vulgaris -eft -spotted salamander, Ambystoma maculatum -axolotl, mud puppy, Ambystoma mexicanum -bullfrog, Rana catesbeiana -tree frog, tree-frog -tailed frog, bell toad, ribbed toad, tailed toad, Ascaphus trui -loggerhead, loggerhead turtle, Caretta caretta -leatherback turtle, leatherback, leathery turtle, Dermochelys coriacea -mud turtle -terrapin -box turtle, box tortoise -banded gecko -common iguana, iguana, Iguana iguana -American chameleon, anole, Anolis carolinensis -whiptail, whiptail lizard -agama -frilled lizard, Chlamydosaurus kingi -alligator lizard -Gila monster, Heloderma suspectum -green lizard, Lacerta viridis -African chameleon, Chamaeleo chamaeleon -Komodo dragon, Komodo lizard, dragon lizard, giant lizard, Varanus komodoensis -African crocodile, Nile crocodile, Crocodylus niloticus -American alligator, Alligator mississipiensis -triceratops -thunder snake, worm snake, Carphophis amoenus -ringneck snake, ring-necked snake, ring snake -hognose snake, puff adder, sand viper -green snake, grass snake -king snake, kingsnake -garter snake, grass snake -water snake -vine snake -night snake, Hypsiglena torquata -boa constrictor, Constrictor constrictor -rock python, rock snake, Python sebae -Indian cobra, Naja naja -green mamba -sea snake -horned viper, cerastes, sand viper, horned asp, Cerastes cornutus -diamondback, diamondback rattlesnake, Crotalus adamanteus -sidewinder, horned rattlesnake, Crotalus cerastes -trilobite -harvestman, daddy longlegs, Phalangium opilio -scorpion -black and gold garden spider, Argiope aurantia -barn spider, Araneus cavaticus -garden spider, Aranea diademata -black widow, Latrodectus mactans -tarantula -wolf spider, hunting spider -tick -centipede -black grouse -ptarmigan -ruffed grouse, partridge, Bonasa umbellus -prairie chicken, prairie grouse, prairie fowl -peacock -quail -partridge -African grey, African gray, Psittacus erithacus -macaw -sulphur-crested cockatoo, Kakatoe galerita, Cacatua galerita -lorikeet -coucal -bee eater -hornbill -hummingbird -jacamar -toucan -drake -red-breasted merganser, Mergus serrator -goose -black swan, Cygnus atratus -tusker -echidna, spiny anteater, anteater -platypus, duckbill, duckbilled platypus, duck-billed platypus, Ornithorhynchus anatinus -wallaby, brush kangaroo -koala, koala bear, kangaroo bear, native bear, Phascolarctos cinereus -wombat -jellyfish -sea anemone, anemone -brain coral -flatworm, platyhelminth -nematode, nematode worm, roundworm -conch -snail -slug -sea slug, nudibranch -chiton, coat-of-mail shell, sea cradle, polyplacophore -chambered nautilus, pearly nautilus, nautilus -Dungeness crab, Cancer magister -rock crab, Cancer irroratus -fiddler crab -king crab, Alaska crab, Alaskan king crab, Alaska king crab, Paralithodes camtschatica -American lobster, Northern lobster, Maine lobster, Homarus americanus -spiny lobster, langouste, rock lobster, crawfish, crayfish, sea crawfish -crayfish, crawfish, crawdad, crawdaddy -hermit crab -isopod -white stork, Ciconia ciconia -black stork, Ciconia nigra -spoonbill -flamingo -little blue heron, Egretta caerulea -American egret, great white heron, Egretta albus -bittern -crane -limpkin, Aramus pictus -European gallinule, Porphyrio porphyrio -American coot, marsh hen, mud hen, water hen, Fulica americana -bustard -ruddy turnstone, Arenaria interpres -red-backed sandpiper, dunlin, Erolia alpina -redshank, Tringa totanus -dowitcher -oystercatcher, oyster catcher -pelican -king penguin, Aptenodytes patagonica -albatross, mollymawk -grey whale, gray whale, devilfish, Eschrichtius gibbosus, Eschrichtius robustus -killer whale, killer, orca, grampus, sea wolf, Orcinus orca -dugong, Dugong dugon -sea lion -Chihuahua -Japanese spaniel -Maltese dog, Maltese terrier, Maltese -Pekinese, Pekingese, Peke -Shih-Tzu -Blenheim spaniel -papillon -toy terrier -Rhodesian ridgeback -Afghan hound, Afghan -basset, basset hound -beagle -bloodhound, sleuthhound -bluetick -black-and-tan coonhound -Walker hound, Walker foxhound -English foxhound -redbone -borzoi, Russian wolfhound -Irish wolfhound -Italian greyhound -whippet -Ibizan hound, Ibizan Podenco -Norwegian elkhound, elkhound -otterhound, otter hound -Saluki, gazelle hound -Scottish deerhound, deerhound -Weimaraner -Staffordshire bullterrier, Staffordshire bull terrier -American Staffordshire terrier, Staffordshire terrier, American pit bull terrier, pit bull terrier -Bedlington terrier -Border terrier -Kerry blue terrier -Irish terrier -Norfolk terrier -Norwich terrier -Yorkshire terrier -wire-haired fox terrier -Lakeland terrier -Sealyham terrier, Sealyham -Airedale, Airedale terrier -cairn, cairn terrier -Australian terrier -Dandie Dinmont, Dandie Dinmont terrier -Boston bull, Boston terrier -miniature schnauzer -giant schnauzer -standard schnauzer -Scotch terrier, Scottish terrier, Scottie -Tibetan terrier, chrysanthemum dog -silky terrier, Sydney silky -soft-coated wheaten terrier -West Highland white terrier -Lhasa, Lhasa apso -flat-coated retriever -curly-coated retriever -golden retriever -Labrador retriever -Chesapeake Bay retriever -German short-haired pointer -vizsla, Hungarian pointer -English setter -Irish setter, red setter -Gordon setter -Brittany spaniel -clumber, clumber spaniel -English springer, English springer spaniel -Welsh springer spaniel -cocker spaniel, English cocker spaniel, cocker -Sussex spaniel -Irish water spaniel -kuvasz -schipperke -groenendael -malinois -briard -kelpie -komondor -Old English sheepdog, bobtail -Shetland sheepdog, Shetland sheep dog, Shetland -collie -Border collie -Bouvier des Flandres, Bouviers des Flandres -Rottweiler -German shepherd, German shepherd dog, German police dog, alsatian -Doberman, Doberman pinscher -miniature pinscher -Greater Swiss Mountain dog -Bernese mountain dog -Appenzeller -EntleBucher -boxer -bull mastiff -Tibetan mastiff -French bulldog -Great Dane -Saint Bernard, St Bernard -Eskimo dog, husky -malamute, malemute, Alaskan malamute -Siberian husky -dalmatian, coach dog, carriage dog -affenpinscher, monkey pinscher, monkey dog -basenji -pug, pug-dog -Leonberg -Newfoundland, Newfoundland dog -Great Pyrenees -Samoyed, Samoyede -Pomeranian -chow, chow chow -keeshond -Brabancon griffon -Pembroke, Pembroke Welsh corgi -Cardigan, Cardigan Welsh corgi -toy poodle -miniature poodle -standard poodle -Mexican hairless -timber wolf, grey wolf, gray wolf, Canis lupus -white wolf, Arctic wolf, Canis lupus tundrarum -red wolf, maned wolf, Canis rufus, Canis niger -coyote, prairie wolf, brush wolf, Canis latrans -dingo, warrigal, warragal, Canis dingo -dhole, Cuon alpinus -African hunting dog, hyena dog, Cape hunting dog, Lycaon pictus -hyena, hyaena -red fox, Vulpes vulpes -kit fox, Vulpes macrotis -Arctic fox, white fox, Alopex lagopus -grey fox, gray fox, Urocyon cinereoargenteus -tabby, tabby cat -tiger cat -Persian cat -Siamese cat, Siamese -Egyptian cat -cougar, puma, catamount, mountain lion, painter, panther, Felis concolor -lynx, catamount -leopard, Panthera pardus -snow leopard, ounce, Panthera uncia -jaguar, panther, Panthera onca, Felis onca -lion, king of beasts, Panthera leo -tiger, Panthera tigris -cheetah, chetah, Acinonyx jubatus -brown bear, bruin, Ursus arctos -American black bear, black bear, Ursus americanus, Euarctos americanus -ice bear, polar bear, Ursus Maritimus, Thalarctos maritimus -sloth bear, Melursus ursinus, Ursus ursinus -mongoose -meerkat, mierkat -tiger beetle -ladybug, ladybeetle, lady beetle, ladybird, ladybird beetle -ground beetle, carabid beetle -long-horned beetle, longicorn, longicorn beetle -leaf beetle, chrysomelid -dung beetle -rhinoceros beetle -weevil -fly -bee -ant, emmet, pismire -grasshopper, hopper -cricket -walking stick, walkingstick, stick insect -cockroach, roach -mantis, mantid -cicada, cicala -leafhopper -lacewing, lacewing fly -dragonfly, darning needle, devil's darning needle, sewing needle, snake feeder, snake doctor, mosquito hawk, skeeter hawk -damselfly -admiral -ringlet, ringlet butterfly -monarch, monarch butterfly, milkweed butterfly, Danaus plexippus -cabbage butterfly -sulphur butterfly, sulfur butterfly -lycaenid, lycaenid butterfly -starfish, sea star -sea urchin -sea cucumber, holothurian -wood rabbit, cottontail, cottontail rabbit -hare -Angora, Angora rabbit -hamster -porcupine, hedgehog -fox squirrel, eastern fox squirrel, Sciurus niger -marmot -beaver -guinea pig, Cavia cobaya -sorrel -zebra -hog, pig, grunter, squealer, Sus scrofa -wild boar, boar, Sus scrofa -warthog -hippopotamus, hippo, river horse, Hippopotamus amphibius -ox -water buffalo, water ox, Asiatic buffalo, Bubalus bubalis -bison -ram, tup -bighorn, bighorn sheep, cimarron, Rocky Mountain bighorn, Rocky Mountain sheep, Ovis canadensis -ibex, Capra ibex -hartebeest -impala, Aepyceros melampus -gazelle -Arabian camel, dromedary, Camelus dromedarius -llama -weasel -mink -polecat, fitch, foulmart, foumart, Mustela putorius -black-footed ferret, ferret, Mustela nigripes -otter -skunk, polecat, wood pussy -badger -armadillo -three-toed sloth, ai, Bradypus tridactylus -orangutan, orang, orangutang, Pongo pygmaeus -gorilla, Gorilla gorilla -chimpanzee, chimp, Pan troglodytes -gibbon, Hylobates lar -siamang, Hylobates syndactylus, Symphalangus syndactylus -guenon, guenon monkey -patas, hussar monkey, Erythrocebus patas -baboon -macaque -langur -colobus, colobus monkey -proboscis monkey, Nasalis larvatus -marmoset -capuchin, ringtail, Cebus capucinus -howler monkey, howler -titi, titi monkey -spider monkey, Ateles geoffroyi -squirrel monkey, Saimiri sciureus -Madagascar cat, ring-tailed lemur, Lemur catta -indri, indris, Indri indri, Indri brevicaudatus -Indian elephant, Elephas maximus -African elephant, Loxodonta africana -lesser panda, red panda, panda, bear cat, cat bear, Ailurus fulgens -giant panda, panda, panda bear, coon bear, Ailuropoda melanoleuca -barracouta, snoek -eel -coho, cohoe, coho salmon, blue jack, silver salmon, Oncorhynchus kisutch -rock beauty, Holocanthus tricolor -anemone fish -sturgeon -gar, garfish, garpike, billfish, Lepisosteus osseus -lionfish -puffer, pufferfish, blowfish, globefish -abacus -abaya -academic gown, academic robe, judge's robe -accordion, piano accordion, squeeze box -acoustic guitar -aircraft carrier, carrier, flattop, attack aircraft carrier -airliner -airship, dirigible -altar -ambulance -amphibian, amphibious vehicle -analog clock -apiary, bee house -apron -ashcan, trash can, garbage can, wastebin, ash bin, ash-bin, ashbin, dustbin, trash barrel, trash bin -assault rifle, assault gun -backpack, back pack, knapsack, packsack, rucksack, haversack -bakery, bakeshop, bakehouse -balance beam, beam -balloon -ballpoint, ballpoint pen, ballpen, Biro -Band Aid -banjo -bannister, banister, balustrade, balusters, handrail -barbell -barber chair -barbershop -barn -barometer -barrel, cask -barrow, garden cart, lawn cart, wheelbarrow -baseball -basketball -bassinet -bassoon -bathing cap, swimming cap -bath towel -bathtub, bathing tub, bath, tub -beach wagon, station wagon, wagon, estate car, beach waggon, station waggon, waggon -beacon, lighthouse, beacon light, pharos -beaker -bearskin, busby, shako -beer bottle -beer glass -bell cote, bell cot -bib -bicycle-built-for-two, tandem bicycle, tandem -bikini, two-piece -binder, ring-binder -binoculars, field glasses, opera glasses -birdhouse -boathouse -bobsled, bobsleigh, bob -bolo tie, bolo, bola tie, bola -bonnet, poke bonnet -bookcase -bookshop, bookstore, bookstall -bottlecap -bow -bow tie, bow-tie, bowtie -brass, memorial tablet, plaque -brassiere, bra, bandeau -breakwater, groin, groyne, mole, bulwark, seawall, jetty -breastplate, aegis, egis -broom -bucket, pail -buckle -bulletproof vest -bullet train, bullet -butcher shop, meat market -cab, hack, taxi, taxicab -caldron, cauldron -candle, taper, wax light -cannon -canoe -can opener, tin opener -cardigan -car mirror -carousel, carrousel, merry-go-round, roundabout, whirligig -carpenter's kit, tool kit -carton -car wheel -cash machine, cash dispenser, automated teller machine, automatic teller machine, automated teller, automatic teller, ATM -cassette -cassette player -castle -catamaran -CD player -cello, violoncello -cellular telephone, cellular phone, cellphone, cell, mobile phone -chain -chainlink fence -chain mail, ring mail, mail, chain armor, chain armour, ring armor, ring armour -chain saw, chainsaw -chest -chiffonier, commode -chime, bell, gong -china cabinet, china closet -Christmas stocking -church, church building -cinema, movie theater, movie theatre, movie house, picture palace -cleaver, meat cleaver, chopper -cliff dwelling -cloak -clog, geta, patten, sabot -cocktail shaker -coffee mug -coffeepot -coil, spiral, volute, whorl, helix -combination lock -computer keyboard, keypad -confectionery, confectionary, candy store -container ship, containership, container vessel -convertible -corkscrew, bottle screw -cornet, horn, trumpet, trump -cowboy boot -cowboy hat, ten-gallon hat -cradle -crane -crash helmet -crate -crib, cot -Crock Pot -croquet ball -crutch -cuirass -dam, dike, dyke -desk -desktop computer -dial telephone, dial phone -diaper, nappy, napkin -digital clock -digital watch -dining table, board -dishrag, dishcloth -dishwasher, dish washer, dishwashing machine -disk brake, disc brake -dock, dockage, docking facility -dogsled, dog sled, dog sleigh -dome -doormat, welcome mat -drilling platform, offshore rig -drum, membranophone, tympan -drumstick -dumbbell -Dutch oven -electric fan, blower -electric guitar -electric locomotive -entertainment center -envelope -espresso maker -face powder -feather boa, boa -file, file cabinet, filing cabinet -fireboat -fire engine, fire truck -fire screen, fireguard -flagpole, flagstaff -flute, transverse flute -folding chair -football helmet -forklift -fountain -fountain pen -four-poster -freight car -French horn, horn -frying pan, frypan, skillet -fur coat -garbage truck, dustcart -gasmask, respirator, gas helmet -gas pump, gasoline pump, petrol pump, island dispenser -goblet -go-kart -golf ball -golfcart, golf cart -gondola -gong, tam-tam -gown -grand piano, grand -greenhouse, nursery, glasshouse -grille, radiator grille -grocery store, grocery, food market, market -guillotine -hair slide -hair spray -half track -hammer -hamper -hand blower, blow dryer, blow drier, hair dryer, hair drier -hand-held computer, hand-held microcomputer -handkerchief, hankie, hanky, hankey -hard disc, hard disk, fixed disk -harmonica, mouth organ, harp, mouth harp -harp -harvester, reaper -hatchet -holster -home theater, home theatre -honeycomb -hook, claw -hoopskirt, crinoline -horizontal bar, high bar -horse cart, horse-cart -hourglass -iPod -iron, smoothing iron -jack-o'-lantern -jean, blue jean, denim -jeep, landrover -jersey, T-shirt, tee shirt -jigsaw puzzle -jinrikisha, ricksha, rickshaw -joystick -kimono -knee pad -knot -lab coat, laboratory coat -ladle -lampshade, lamp shade -laptop, laptop computer -lawn mower, mower -lens cap, lens cover -letter opener, paper knife, paperknife -library -lifeboat -lighter, light, igniter, ignitor -limousine, limo -liner, ocean liner -lipstick, lip rouge -Loafer -lotion -loudspeaker, speaker, speaker unit, loudspeaker system, speaker system -loupe, jeweler's loupe -lumbermill, sawmill -magnetic compass -mailbag, postbag -mailbox, letter box -maillot -maillot, tank suit -manhole cover -maraca -marimba, xylophone -mask -matchstick -maypole -maze, labyrinth -measuring cup -medicine chest, medicine cabinet -megalith, megalithic structure -microphone, mike -microwave, microwave oven -military uniform -milk can -minibus -miniskirt, mini -minivan -missile -mitten -mixing bowl -mobile home, manufactured home -Model T -modem -monastery -monitor -moped -mortar -mortarboard -mosque -mosquito net -motor scooter, scooter -mountain bike, all-terrain bike, off-roader -mountain tent -mouse, computer mouse -mousetrap -moving van -muzzle -nail -neck brace -necklace -nipple -notebook, notebook computer -obelisk -oboe, hautboy, hautbois -ocarina, sweet potato -odometer, hodometer, mileometer, milometer -oil filter -organ, pipe organ -oscilloscope, scope, cathode-ray oscilloscope, CRO -overskirt -oxcart -oxygen mask -packet -paddle, boat paddle -paddlewheel, paddle wheel -padlock -paintbrush -pajama, pyjama, pj's, jammies -palace -panpipe, pandean pipe, syrinx -paper towel -parachute, chute -parallel bars, bars -park bench -parking meter -passenger car, coach, carriage -patio, terrace -pay-phone, pay-station -pedestal, plinth, footstall -pencil box, pencil case -pencil sharpener -perfume, essence -Petri dish -photocopier -pick, plectrum, plectron -pickelhaube -picket fence, paling -pickup, pickup truck -pier -piggy bank, penny bank -pill bottle -pillow -ping-pong ball -pinwheel -pirate, pirate ship -pitcher, ewer -plane, carpenter's plane, woodworking plane -planetarium -plastic bag -plate rack -plow, plough -plunger, plumber's helper -Polaroid camera, Polaroid Land camera -pole -police van, police wagon, paddy wagon, patrol wagon, wagon, black Maria -poncho -pool table, billiard table, snooker table -pop bottle, soda bottle -pot, flowerpot -potter's wheel -power drill -prayer rug, prayer mat -printer -prison, prison house -projectile, missile -projector -puck, hockey puck -punching bag, punch bag, punching ball, punchball -purse -quill, quill pen -quilt, comforter, comfort, puff -racer, race car, racing car -racket, racquet -radiator -radio, wireless -radio telescope, radio reflector -rain barrel -recreational vehicle, RV, R.V. -reel -reflex camera -refrigerator, icebox -remote control, remote -restaurant, eating house, eating place, eatery -revolver, six-gun, six-shooter -rifle -rocking chair, rocker -rotisserie -rubber eraser, rubber, pencil eraser -rugby ball -rule, ruler -running shoe -safe -safety pin -saltshaker, salt shaker -sandal -sarong -sax, saxophone -scabbard -scale, weighing machine -school bus -schooner -scoreboard -screen, CRT screen -screw -screwdriver -seat belt, seatbelt -sewing machine -shield, buckler -shoe shop, shoe-shop, shoe store -shoji -shopping basket -shopping cart -shovel -shower cap -shower curtain -ski -ski mask -sleeping bag -slide rule, slipstick -sliding door -slot, one-armed bandit -snorkel -snowmobile -snowplow, snowplough -soap dispenser -soccer ball -sock -solar dish, solar collector, solar furnace -sombrero -soup bowl -space bar -space heater -space shuttle -spatula -speedboat -spider web, spider's web -spindle -sports car, sport car -spotlight, spot -stage -steam locomotive -steel arch bridge -steel drum -stethoscope -stole -stone wall -stopwatch, stop watch -stove -strainer -streetcar, tram, tramcar, trolley, trolley car -stretcher -studio couch, day bed -stupa, tope -submarine, pigboat, sub, U-boat -suit, suit of clothes -sundial -sunglass -sunglasses, dark glasses, shades -sunscreen, sunblock, sun blocker -suspension bridge -swab, swob, mop -sweatshirt -swimming trunks, bathing trunks -swing -switch, electric switch, electrical switch -syringe -table lamp -tank, army tank, armored combat vehicle, armoured combat vehicle -tape player -teapot -teddy, teddy bear -television, television system -tennis ball -thatch, thatched roof -theater curtain, theatre curtain -thimble -thresher, thrasher, threshing machine -throne -tile roof -toaster -tobacco shop, tobacconist shop, tobacconist -toilet seat -torch -totem pole -tow truck, tow car, wrecker -toyshop -tractor -trailer truck, tractor trailer, trucking rig, rig, articulated lorry, semi -tray -trench coat -tricycle, trike, velocipede -trimaran -tripod -triumphal arch -trolleybus, trolley coach, trackless trolley -trombone -tub, vat -turnstile -typewriter keyboard -umbrella -unicycle, monocycle -upright, upright piano -vacuum, vacuum cleaner -vase -vault -velvet -vending machine -vestment -viaduct -violin, fiddle -volleyball -waffle iron -wall clock -wallet, billfold, notecase, pocketbook -wardrobe, closet, press -warplane, military plane -washbasin, handbasin, washbowl, lavabo, wash-hand basin -washer, automatic washer, washing machine -water bottle -water jug -water tower -whiskey jug -whistle -wig -window screen -window shade -Windsor tie -wine bottle -wing -wok -wooden spoon -wool, woolen, woollen -worm fence, snake fence, snake-rail fence, Virginia fence -wreck -yawl -yurt -web site, website, internet site, site -comic book -crossword puzzle, crossword -street sign -traffic light, traffic signal, stoplight -book jacket, dust cover, dust jacket, dust wrapper -menu -plate -guacamole -consomme -hot pot, hotpot -trifle -ice cream, icecream -ice lolly, lolly, lollipop, popsicle -French loaf -bagel, beigel -pretzel -cheeseburger -hotdog, hot dog, red hot -mashed potato -head cabbage -broccoli -cauliflower -zucchini, courgette -spaghetti squash -acorn squash -butternut squash -cucumber, cuke -artichoke, globe artichoke -bell pepper -cardoon -mushroom -Granny Smith -strawberry -orange -lemon -fig -pineapple, ananas -banana -jackfruit, jak, jack -custard apple -pomegranate -hay -carbonara -chocolate sauce, chocolate syrup -dough -meat loaf, meatloaf -pizza, pizza pie -potpie -burrito -red wine -espresso -cup -eggnog -alp -bubble -cliff, drop, drop-off -coral reef -geyser -lakeside, lakeshore -promontory, headland, head, foreland -sandbar, sand bar -seashore, coast, seacoast, sea-coast -valley, vale -volcano -ballplayer, baseball player -groom, bridegroom -scuba diver -rapeseed -daisy -yellow lady's slipper, yellow lady-slipper, Cypripedium calceolus, Cypripedium parviflorum -corn -acorn -hip, rose hip, rosehip -buckeye, horse chestnut, conker -coral fungus -agaric -gyromitra -stinkhorn, carrion fungus -earthstar -hen-of-the-woods, hen of the woods, Polyporus frondosus, Grifola frondosa -bolete -ear, spike, capitulum -toilet tissue, toilet paper, bathroom tissue diff --git a/hub_module/modules/image/classification/efficientnetb7_imagenet/layers.py b/hub_module/modules/image/classification/efficientnetb7_imagenet/layers.py deleted file mode 100644 index 00f792b5..00000000 --- a/hub_module/modules/image/classification/efficientnetb7_imagenet/layers.py +++ /dev/null @@ -1,248 +0,0 @@ -# -*- coding:utf-8 -*- -# copyright (c) 2020 PaddlePaddle Authors. All Rights Reserve. -# -# Licensed under the Apache License, Version 2.0 (the "License"); -# you may not use this file except in compliance with the License. -# You may obtain a copy of the License at -# -# http://www.apache.org/licenses/LICENSE-2.0 -# -# Unless required by applicable law or agreed to in writing, software -# distributed under the License is distributed on an "AS IS" BASIS, -# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. -# See the License for the specific language governing permissions and -# limitations under the License. - -from __future__ import absolute_import -from __future__ import division -from __future__ import print_function - -import math -import warnings - -import paddle.fluid as fluid - - -def initial_type(name, - input, - op_type, - fan_out, - init="google", - use_bias=False, - filter_size=0, - stddev=0.02): - if init == "kaiming": - if op_type == 'conv': - fan_in = input.shape[1] * filter_size * filter_size - elif op_type == 'deconv': - fan_in = fan_out * filter_size * filter_size - else: - if len(input.shape) > 2: - fan_in = input.shape[1] * input.shape[2] * input.shape[3] - else: - fan_in = input.shape[1] - bound = 1 / math.sqrt(fan_in) - param_attr = fluid.ParamAttr( - name=name + "_weights", - initializer=fluid.initializer.Uniform(low=-bound, high=bound)) - if use_bias == True: - bias_attr = fluid.ParamAttr( - name=name + '_offset', - initializer=fluid.initializer.Uniform(low=-bound, high=bound)) - else: - bias_attr = False - elif init == 'google': - n = filter_size * filter_size * fan_out - param_attr = fluid.ParamAttr( - name=name + "_weights", - initializer=fluid.initializer.NormalInitializer( - loc=0.0, scale=math.sqrt(2.0 / n))) - if use_bias == True: - bias_attr = fluid.ParamAttr( - name=name + "_offset", - initializer=fluid.initializer.Constant(0.0)) - else: - bias_attr = False - - else: - param_attr = fluid.ParamAttr( - name=name + "_weights", - initializer=fluid.initializer.NormalInitializer( - loc=0.0, scale=stddev)) - if use_bias == True: - bias_attr = fluid.ParamAttr( - name=name + "_offset", - initializer=fluid.initializer.Constant(0.0)) - else: - bias_attr = False - return param_attr, bias_attr - - -def cal_padding(img_size, stride, filter_size, dilation=1): - """Calculate padding size.""" - if img_size % stride == 0: - out_size = max(filter_size - stride, 0) - else: - out_size = max(filter_size - (img_size % stride), 0) - return out_size // 2, out_size - out_size // 2 - - -def init_batch_norm_layer(name="batch_norm"): - param_attr = fluid.ParamAttr( - name=name + '_scale', initializer=fluid.initializer.Constant(1.0)) - bias_attr = fluid.ParamAttr( - name=name + '_offset', - initializer=fluid.initializer.Constant(value=0.0)) - return param_attr, bias_attr - - -def init_fc_layer(fout, name='fc'): - n = fout # fan-out - init_range = 1.0 / math.sqrt(n) - - param_attr = fluid.ParamAttr( - name=name + '_weights', - initializer=fluid.initializer.UniformInitializer( - low=-init_range, high=init_range)) - bias_attr = fluid.ParamAttr( - name=name + '_offset', - initializer=fluid.initializer.Constant(value=0.0)) - return param_attr, bias_attr - - -def norm_layer(input, norm_type='batch_norm', name=None): - if norm_type == 'batch_norm': - param_attr = fluid.ParamAttr( - name=name + '_weights', initializer=fluid.initializer.Constant(1.0)) - bias_attr = fluid.ParamAttr( - name=name + '_offset', - initializer=fluid.initializer.Constant(value=0.0)) - return fluid.layers.batch_norm( - input, - param_attr=param_attr, - bias_attr=bias_attr, - moving_mean_name=name + '_mean', - moving_variance_name=name + '_variance') - - elif norm_type == 'instance_norm': - helper = fluid.layer_helper.LayerHelper("instance_norm", **locals()) - dtype = helper.input_dtype() - epsilon = 1e-5 - mean = fluid.layers.reduce_mean(input, dim=[2, 3], keep_dim=True) - var = fluid.layers.reduce_mean( - fluid.layers.square(input - mean), dim=[2, 3], keep_dim=True) - if name is not None: - scale_name = name + "_scale" - offset_name = name + "_offset" - scale_param = fluid.ParamAttr( - name=scale_name, - initializer=fluid.initializer.Constant(1.0), - trainable=True) - offset_param = fluid.ParamAttr( - name=offset_name, - initializer=fluid.initializer.Constant(0.0), - trainable=True) - scale = helper.create_parameter( - attr=scale_param, shape=input.shape[1:2], dtype=dtype) - offset = helper.create_parameter( - attr=offset_param, shape=input.shape[1:2], dtype=dtype) - - tmp = fluid.layers.elementwise_mul(x=(input - mean), y=scale, axis=1) - tmp = tmp / fluid.layers.sqrt(var + epsilon) - tmp = fluid.layers.elementwise_add(tmp, offset, axis=1) - return tmp - else: - raise NotImplementedError("norm tyoe: [%s] is not support" % norm_type) - - -def conv2d(input, - num_filters=64, - filter_size=7, - stride=1, - stddev=0.02, - padding=0, - groups=None, - name="conv2d", - norm=None, - act=None, - relufactor=0.0, - use_bias=False, - padding_type=None, - initial="normal", - use_cudnn=True): - - if padding != 0 and padding_type != None: - warnings.warn( - 'padding value and padding type are set in the same time, and the final padding width and padding height are computed by padding_type' - ) - - param_attr, bias_attr = initial_type( - name=name, - input=input, - op_type='conv', - fan_out=num_filters, - init=initial, - use_bias=use_bias, - filter_size=filter_size, - stddev=stddev) - - def get_padding(filter_size, stride=1, dilation=1): - padding = ((stride - 1) + dilation * (filter_size - 1)) // 2 - return padding - - need_crop = False - if padding_type == "SAME": - top_padding, bottom_padding = cal_padding(input.shape[2], stride, - filter_size) - left_padding, right_padding = cal_padding(input.shape[2], stride, - filter_size) - height_padding = bottom_padding - width_padding = right_padding - if top_padding != bottom_padding or left_padding != right_padding: - height_padding = top_padding + stride - width_padding = left_padding + stride - need_crop = True - padding = [height_padding, width_padding] - elif padding_type == "VALID": - height_padding = 0 - width_padding = 0 - padding = [height_padding, width_padding] - elif padding_type == "DYNAMIC": - padding = get_padding(filter_size, stride) - else: - padding = padding - - conv = fluid.layers.conv2d( - input, - num_filters, - filter_size, - groups=groups, - name=name, - stride=stride, - padding=padding, - use_cudnn=use_cudnn, - param_attr=param_attr, - bias_attr=bias_attr) - - if need_crop: - conv = conv[:, :, 1:, 1:] - - if norm is not None: - conv = norm_layer(input=conv, norm_type=norm, name=name + "_norm") - if act == 'relu': - conv = fluid.layers.relu(conv, name=name + '_relu') - elif act == 'leaky_relu': - conv = fluid.layers.leaky_relu( - conv, alpha=relufactor, name=name + '_leaky_relu') - elif act == 'tanh': - conv = fluid.layers.tanh(conv, name=name + '_tanh') - elif act == 'sigmoid': - conv = fluid.layers.sigmoid(conv, name=name + '_sigmoid') - elif act == 'swish': - conv = fluid.layers.swish(conv, name=name + '_swish') - elif act == None: - conv = conv - else: - raise NotImplementedError("activation: [%s] is not support" % act) - - return conv diff --git a/hub_module/modules/image/classification/efficientnetb7_imagenet/module.py b/hub_module/modules/image/classification/efficientnetb7_imagenet/module.py index 4aca570b..9158dc52 100644 --- a/hub_module/modules/image/classification/efficientnetb7_imagenet/module.py +++ b/hub_module/modules/image/classification/efficientnetb7_imagenet/module.py @@ -1,4 +1,3 @@ -# -*- coding:utf-8 -*- # copyright (c) 2020 PaddlePaddle Authors. All Rights Reserve. # # Licensed under the Apache License, Version 2.0 (the "License"); @@ -13,312 +12,747 @@ # See the License for the specific language governing permissions and # limitations under the License. -from __future__ import absolute_import -from __future__ import division - -import ast -import argparse import os - -import numpy as np -import paddle.fluid as fluid -import paddlehub as hub -from paddle.fluid.core import PaddleTensor, AnalysisConfig, create_paddle_predictor -from paddlehub.module.module import moduleinfo, runnable, serving -from paddlehub.common.paddle_helper import add_vars_prefix - -from efficientnetb7_imagenet.processor import postprocess, base64_to_cv2 -from efficientnetb7_imagenet.data_feed import reader -from efficientnetb7_imagenet.efficientnet import EfficientNetB7 - - -@moduleinfo( - name="efficientnetb7_imagenet", - type="CV/image_classification", - author="paddlepaddle", - author_email="paddle-dev@baidu.com", - summary= - "EfficientNetB7 is a image classfication model, this module is trained with imagenet datasets.", - version="1.1.0") -class EfficientNetB7ImageNet(hub.Module): - def _initialize(self): - self.default_pretrained_model_path = os.path.join( - self.directory, "efficientnetb7_imagenet_infer_model") - label_file = os.path.join(self.directory, "label_list.txt") - with open(label_file, 'r', encoding='utf-8') as file: - self.label_list = file.read().split("\n")[:-1] - self.classification = self.classify - self._set_config() - - def get_expected_image_width(self): - return 224 - - def get_expected_image_height(self): - return 224 - - def get_pretrained_images_mean(self): - im_mean = np.array([0.485, 0.456, 0.406]).reshape(1, 3) - return im_mean - - def get_pretrained_images_std(self): - im_std = np.array([0.229, 0.224, 0.225]).reshape(1, 3) - return im_std - - def _set_config(self): - """ - predictor config setting +import math +import collections +import re +import copy + +import paddle +from paddle import ParamAttr +import paddle.nn as nn +import paddle.nn.functional as F +from paddle.nn import Conv2d, BatchNorm, Linear, Dropout +from paddle.nn import AdaptiveAvgPool2d, MaxPool2d, AvgPool2d +from paddlehub.module.module import moduleinfo +from paddlehub.module.cv_module import ImageClassifierModule + +GlobalParams = collections.namedtuple('GlobalParams', [ + 'batch_norm_momentum', + 'batch_norm_epsilon', + 'dropout_rate', + 'num_classes', + 'width_coefficient', + 'depth_coefficient', + 'depth_divisor', + 'min_depth', + 'drop_connect_rate', +]) + +BlockArgs = collections.namedtuple( + 'BlockArgs', + ['kernel_size', 'num_repeat', 'input_filters', 'output_filters', 'expand_ratio', 'id_skip', 'stride', 'se_ratio']) + +GlobalParams.__new__.__defaults__ = (None, ) * len(GlobalParams._fields) +BlockArgs.__new__.__defaults__ = (None, ) * len(BlockArgs._fields) + + +def efficientnet_params(model_name: str): + """ Map EfficientNet model name to parameter coefficients. """ + params_dict = { + # Coefficients: width,depth,resolution,dropout + 'efficientnet-b7': (2.0, 3.1, 600, 0.5) + } + return params_dict[model_name] + + +def efficientnet(width_coefficient: float = None, + depth_coefficient: float = None, + dropout_rate: float = 0.2, + drop_connect_rate: float = 0.2): + """ Get block arguments according to parameter and coefficients. """ + blocks_args = [ + 'r1_k3_s11_e1_i32_o16_se0.25', + 'r2_k3_s22_e6_i16_o24_se0.25', + 'r2_k5_s22_e6_i24_o40_se0.25', + 'r3_k3_s22_e6_i40_o80_se0.25', + 'r3_k5_s11_e6_i80_o112_se0.25', + 'r4_k5_s22_e6_i112_o192_se0.25', + 'r1_k3_s11_e6_i192_o320_se0.25', + ] + blocks_args = BlockDecoder.decode(blocks_args) + + global_params = GlobalParams(batch_norm_momentum=0.99, + batch_norm_epsilon=1e-3, + dropout_rate=dropout_rate, + drop_connect_rate=drop_connect_rate, + num_classes=1000, + width_coefficient=width_coefficient, + depth_coefficient=depth_coefficient, + depth_divisor=8, + min_depth=None) + + return blocks_args, global_params + + +def get_model_params(model_name: str, override_params: dict): + """ Get the block args and global params for a given model """ + if model_name.startswith('efficientnet'): + w, d, _, p = efficientnet_params(model_name) + blocks_args, global_params = efficientnet(width_coefficient=w, depth_coefficient=d, dropout_rate=p) + else: + raise NotImplementedError('model name is not pre-defined: %s' % model_name) + if override_params: + global_params = global_params._replace(**override_params) + return blocks_args, global_params + + +def round_filters(filters: int, global_params: dict): + """ Calculate and round number of filters based on depth multiplier. """ + multiplier = global_params.width_coefficient + if not multiplier: + return filters + divisor = global_params.depth_divisor + min_depth = global_params.min_depth + filters *= multiplier + min_depth = min_depth or divisor + new_filters = max(min_depth, int(filters + divisor / 2) // divisor * divisor) + if new_filters < 0.9 * filters: # prevent rounding by more than 10% + new_filters += divisor + return int(new_filters) + + +def round_repeats(repeats: int, global_params: dict): + """ Round number of filters based on depth multiplier. """ + multiplier = global_params.depth_coefficient + if not multiplier: + return repeats + return int(math.ceil(multiplier * repeats)) + + +class BlockDecoder(object): + """ + Block Decoder, straight from the official TensorFlow repository. + """ + @staticmethod + def _decode_block_string(block_string: str): + """ Gets a block through a string notation of arguments. """ + assert isinstance(block_string, str) + + ops = block_string.split('_') + options = {} + for op in ops: + splits = re.split(r'(\d.*)', op) + if len(splits) >= 2: + key, value = splits[:2] + options[key] = value + + # Check stride + cond_1 = ('s' in options and len(options['s']) == 1) + cond_2 = ((len(options['s']) == 2) and (options['s'][0] == options['s'][1])) + assert (cond_1 or cond_2) + + return BlockArgs(kernel_size=int(options['k']), + num_repeat=int(options['r']), + input_filters=int(options['i']), + output_filters=int(options['o']), + expand_ratio=int(options['e']), + id_skip=('noskip' not in block_string), + se_ratio=float(options['se']) if 'se' in options else None, + stride=[int(options['s'][0])]) + + @staticmethod + def _encode_block_string(block): + """Encodes a block to a string.""" + args = [ + 'r%d' % block.num_repeat, + 'k%d' % block.kernel_size, + 's%d%d' % (block.strides[0], block.strides[1]), + 'e%s' % block.expand_ratio, + 'i%d' % block.input_filters, + 'o%d' % block.output_filters + ] + if 0 < block.se_ratio <= 1: + args.append('se%s' % block.se_ratio) + if block.id_skip is False: + args.append('noskip') + return '_'.join(args) + + @staticmethod + def decode(string_list: list): """ - cpu_config = AnalysisConfig(self.default_pretrained_model_path) - cpu_config.disable_glog_info() - cpu_config.disable_gpu() - self.cpu_predictor = create_paddle_predictor(cpu_config) - - try: - _places = os.environ["CUDA_VISIBLE_DEVICES"] - int(_places[0]) - use_gpu = True - except: - use_gpu = False - if use_gpu: - gpu_config = AnalysisConfig(self.default_pretrained_model_path) - gpu_config.disable_glog_info() - gpu_config.enable_use_gpu( - memory_pool_init_size_mb=1000, device_id=0) - self.gpu_predictor = create_paddle_predictor(gpu_config) - - def context(self, - trainable=True, - pretrained=True, - override_params=None, - phase='train'): - """context for transfer learning. - - Args: - trainable (bool): Set parameters in program to be trainable. - pretrained (bool) : Whether to load pretrained model. - - Returns: - inputs (dict): key is 'image', corresponding vaule is image tensor. - outputs (dict): key is : - 'classification', corresponding value is the result of classification. - 'feature_map', corresponding value is the result of the layer before the fully connected layer. - context_prog (fluid.Program): program for transfer learning. - """ - if phase in ["dev", "test", "predict", "eval"]: - is_test = True - elif phase in ["train"]: - is_test = False - else: - raise ValueError( - "Phase %s is error, which must be one of train, dev, test, eval and predict." - % phase) - - context_prog = fluid.Program() - startup_prog = fluid.Program() - with fluid.program_guard(context_prog, startup_prog): - with fluid.unique_name.guard(): - image = fluid.layers.data( - name="image", shape=[3, 224, 224], dtype="float32") - efficientnet_b7 = EfficientNetB7( - override_params=override_params) - output, feature_map = efficientnet_b7.net( - input=image, - class_dim=len(self.label_list), - is_test=is_test) - - name_prefix = '@HUB_{}@'.format(self.name) - inputs = {'image': name_prefix + image.name} - outputs = { - 'classification': name_prefix + output.name, - 'feature_map': name_prefix + feature_map.name - } - add_vars_prefix(context_prog, name_prefix) - add_vars_prefix(startup_prog, name_prefix) - global_vars = context_prog.global_block().vars - inputs = { - key: global_vars[value] - for key, value in inputs.items() - } - outputs = { - key: global_vars[value] - for key, value in outputs.items() - } - - place = fluid.CPUPlace() - exe = fluid.Executor(place) - # pretrained - if pretrained: - - def _if_exist(var): - b = os.path.exists( - os.path.join(self.default_pretrained_model_path, - var.name)) - return b - - fluid.io.load_vars( - exe, - self.default_pretrained_model_path, - context_prog, - predicate=_if_exist) - else: - exe.run(startup_prog) - # trainable - for param in context_prog.global_block().iter_parameters(): - param.trainable = trainable - return inputs, outputs, context_prog - - def classify(self, - images=None, - paths=None, - batch_size=1, - use_gpu=False, - top_k=1): - """ - API for image classification. - - Args: - images (list[numpy.ndarray]): data of images, shape of each is [H, W, C], color space must be BGR. - paths (list[str]): The paths of images. - batch_size (int): batch size. - use_gpu (bool): Whether to use gpu. - top_k (int): Return top k results. + Decode a list of string notations to specify blocks in the network. - Returns: - res (list[dict]): The classfication results. + string_list: list of strings, each string is a notation of block + return + list of BlockArgs namedtuples of block args """ - if use_gpu: - try: - _places = os.environ["CUDA_VISIBLE_DEVICES"] - int(_places[0]) - except: - raise RuntimeError( - "Environment Variable CUDA_VISIBLE_DEVICES is not set correctly. If you wanna use gpu, please set CUDA_VISIBLE_DEVICES as cuda_device_id." - ) - - all_data = list() - for yield_data in reader(images, paths): - all_data.append(yield_data) - - total_num = len(all_data) - loop_num = int(np.ceil(total_num / batch_size)) - - res = list() - for iter_id in range(loop_num): - batch_data = list() - handle_id = iter_id * batch_size - for image_id in range(batch_size): - try: - batch_data.append(all_data[handle_id + image_id]) - except: - pass - # feed batch image - batch_image = np.array([data['image'] for data in batch_data]) - batch_image = PaddleTensor(batch_image.copy()) - predictor_output = self.gpu_predictor.run([ - batch_image - ]) if use_gpu else self.cpu_predictor.run([batch_image]) - out = postprocess( - data_out=predictor_output[0].as_ndarray(), - label_list=self.label_list, - top_k=top_k) - res += out - return res - - def save_inference_model(self, - dirname, - model_filename=None, - params_filename=None, - combined=True): - if combined: - model_filename = "__model__" if not model_filename else model_filename - params_filename = "__params__" if not params_filename else params_filename - place = fluid.CPUPlace() - exe = fluid.Executor(place) - - program, feeded_var_names, target_vars = fluid.io.load_inference_model( - dirname=self.default_pretrained_model_path, executor=exe) - - fluid.io.save_inference_model( - dirname=dirname, - main_program=program, - executor=exe, - feeded_var_names=feeded_var_names, - target_vars=target_vars, - model_filename=model_filename, - params_filename=params_filename) - - @serving - def serving_method(self, images, **kwargs): + assert isinstance(string_list, list) + blocks_args = [] + for block_string in string_list: + blocks_args.append(BlockDecoder._decode_block_string(block_string)) + return blocks_args + + @staticmethod + def encode(blocks_args: list): """ - Run as a service. - """ - images_decode = [base64_to_cv2(image) for image in images] - results = self.classify(images=images_decode, **kwargs) - return results + Encodes a list of BlockArgs to a list of strings. - @runnable - def run_cmd(self, argvs): - """ - Run as a command. + :param blocks_args: a list of BlockArgs namedtuples of block args + :return: a list of strings, each string is a notation of block """ - self.parser = argparse.ArgumentParser( - description="Run the {} module.".format(self.name), - prog='hub run {}'.format(self.name), - usage='%(prog)s', - add_help=True) - self.arg_input_group = self.parser.add_argument_group( - title="Input options", description="Input data. Required") - self.arg_config_group = self.parser.add_argument_group( - title="Config options", - description= - "Run configuration for controlling module behavior, not required.") - self.add_module_config_arg() - self.add_module_input_arg() - args = self.parser.parse_args(argvs) - results = self.classify( - paths=[args.input_path], - batch_size=args.batch_size, - use_gpu=args.use_gpu) - return results - - def add_module_config_arg(self): - """ - Add the command config options. - """ - self.arg_config_group.add_argument( - '--use_gpu', - type=ast.literal_eval, - default=False, - help="whether use GPU or not.") - self.arg_config_group.add_argument( - '--batch_size', - type=ast.literal_eval, - default=1, - help="batch size.") - self.arg_config_group.add_argument( - '--top_k', - type=ast.literal_eval, - default=1, - help="Return top k results.") - - def add_module_input_arg(self): - """ - Add the command input options. - """ - self.arg_input_group.add_argument( - '--input_path', type=str, help="path to image.") - - -if __name__ == '__main__': - b7 = EfficientNetB7ImageNet() - b7.context() - import cv2 - test_image = [cv2.imread('dog.jpeg')] - res = b7.classification(images=test_image) - print(res) - res = b7.classification(paths=['dog.jpeg']) - print(res) - res = b7.classification(images=test_image) - print(res) - res = b7.classify(images=test_image) - print(res) + block_strings = [] + for block in blocks_args: + block_strings.append(BlockDecoder._encode_block_string(block)) + return block_strings + + +def initial_type(name: str, use_bias: bool = False): + param_attr = ParamAttr(name=name + "_weights") + if use_bias: + bias_attr = ParamAttr(name=name + "_offset") + else: + bias_attr = False + return param_attr, bias_attr + + +def init_batch_norm_layer(name: str = "batch_norm"): + param_attr = ParamAttr(name=name + "_scale") + bias_attr = ParamAttr(name=name + "_offset") + return param_attr, bias_attr + + +def init_fc_layer(name: str = "fc"): + param_attr = ParamAttr(name=name + "_weights") + bias_attr = ParamAttr(name=name + "_offset") + return param_attr, bias_attr + + +def cal_padding(img_size: int, stride: int, filter_size: int, dilation: int = 1): + """Calculate padding size.""" + if img_size % stride == 0: + out_size = max(filter_size - stride, 0) + else: + out_size = max(filter_size - (img_size % stride), 0) + return out_size // 2, out_size - out_size // 2 + + +inp_shape = {"b7": [600, 300, 300, 150, 75, 38, 38, 19]} + + +def _drop_connect(inputs: paddle.Tensor, prob: float, is_test: bool): + """Drop input connection""" + if is_test: + return inputs + keep_prob = 1.0 - prob + inputs_shape = paddle.shape(inputs) + random_tensor = keep_prob + paddle.rand(shape=[inputs_shape[0], 1, 1, 1]) + binary_tensor = paddle.floor(random_tensor) + output = inputs / keep_prob * binary_tensor + return output + + +class Conv2ds(nn.Layer): + """Basic conv layer""" + def __init__(self, + input_channels: int, + output_channels: int, + filter_size: int, + stride: int = 1, + padding: int = 0, + groups: int = None, + name: str = "conv2d", + act: str = None, + use_bias: bool = False, + padding_type: str = None, + model_name: str = None, + cur_stage: str = None): + super(Conv2ds, self).__init__() + assert act in [None, "swish", "sigmoid"] + self.act = act + + param_attr, bias_attr = initial_type(name=name, use_bias=use_bias) + + def get_padding(filter_size, stride=1, dilation=1): + padding = ((stride - 1) + dilation * (filter_size - 1)) // 2 + return padding + + inps = 1 if model_name == None and cur_stage == None else inp_shape[model_name][cur_stage] + self.need_crop = False + if padding_type == "SAME": + top_padding, bottom_padding = cal_padding(inps, stride, filter_size) + left_padding, right_padding = cal_padding(inps, stride, filter_size) + height_padding = bottom_padding + width_padding = right_padding + if top_padding != bottom_padding or left_padding != right_padding: + height_padding = top_padding + stride + width_padding = left_padding + stride + self.need_crop = True + padding = [height_padding, width_padding] + elif padding_type == "VALID": + height_padding = 0 + width_padding = 0 + padding = [height_padding, width_padding] + elif padding_type == "DYNAMIC": + padding = get_padding(filter_size, stride) + else: + padding = padding_type + + groups = 1 if groups is None else groups + self._conv = Conv2d(input_channels, + output_channels, + filter_size, + groups=groups, + stride=stride, + padding=padding, + weight_attr=param_attr, + bias_attr=bias_attr) + + def forward(self, inputs: paddle.Tensor): + x = self._conv(inputs) + if self.act == "swish": + x = F.swish(x) + elif self.act == "sigmoid": + x = F.sigmoid(x) + + if self.need_crop: + x = x[:, :, 1:, 1:] + return x + + +class ConvBNLayer(nn.Layer): + """Basic conv bn layer.""" + def __init__(self, + input_channels: int, + filter_size: int, + output_channels: int, + stride: int = 1, + num_groups: int = 1, + padding_type: str = "SAME", + conv_act: str = None, + bn_act: str = "swish", + use_bn: bool = True, + use_bias: bool = False, + name: str = None, + conv_name: str = None, + bn_name: str = None, + model_name: str = None, + cur_stage: str = None): + super(ConvBNLayer, self).__init__() + + self._conv = Conv2ds(input_channels=input_channels, + output_channels=output_channels, + filter_size=filter_size, + stride=stride, + groups=num_groups, + act=conv_act, + padding_type=padding_type, + name=conv_name, + use_bias=use_bias, + model_name=model_name, + cur_stage=cur_stage) + self.use_bn = use_bn + if use_bn is True: + bn_name = name + bn_name + param_attr, bias_attr = init_batch_norm_layer(bn_name) + + self._bn = BatchNorm(num_channels=output_channels, + act=bn_act, + momentum=0.99, + epsilon=0.001, + moving_mean_name=bn_name + "_mean", + moving_variance_name=bn_name + "_variance", + param_attr=param_attr, + bias_attr=bias_attr) + + def forward(self, inputs: paddle.Tensor): + if self.use_bn: + x = self._conv(inputs) + x = self._bn(x) + return x + else: + return self._conv(inputs) + + +class ExpandConvNorm(nn.Layer): + """Expand conv norm layer.""" + def __init__(self, + input_channels: int, + block_args: dict, + padding_type: str, + name: str = None, + model_name: str = None, + cur_stage: str = None): + super(ExpandConvNorm, self).__init__() + + self.oup = block_args.input_filters * block_args.expand_ratio + self.expand_ratio = block_args.expand_ratio + + if self.expand_ratio != 1: + self._conv = ConvBNLayer(input_channels, + 1, + self.oup, + bn_act=None, + padding_type=padding_type, + name=name, + conv_name=name + "_expand_conv", + bn_name="_bn0", + model_name=model_name, + cur_stage=cur_stage) + + def forward(self, inputs: paddle.Tensor): + if self.expand_ratio != 1: + return self._conv(inputs) + else: + return inputs + + +class DepthwiseConvNorm(nn.Layer): + """Depthwise conv norm layer.""" + def __init__(self, + input_channels: int, + block_args: dict, + padding_type: str, + name: str = None, + model_name: str = None, + cur_stage: str = None): + super(DepthwiseConvNorm, self).__init__() + + self.k = block_args.kernel_size + self.s = block_args.stride + if isinstance(self.s, list) or isinstance(self.s, tuple): + self.s = self.s[0] + oup = block_args.input_filters * block_args.expand_ratio + + self._conv = ConvBNLayer(input_channels, + self.k, + oup, + self.s, + num_groups=input_channels, + bn_act=None, + padding_type=padding_type, + name=name, + conv_name=name + "_depthwise_conv", + bn_name="_bn1", + model_name=model_name, + cur_stage=cur_stage) + + def forward(self, inputs: paddle.Tensor): + return self._conv(inputs) + + +class ProjectConvNorm(nn.Layer): + """Projection conv bn layer.""" + def __init__(self, + input_channels: int, + block_args: dict, + padding_type: str, + name: str = None, + model_name: str = None, + cur_stage: str = None): + super(ProjectConvNorm, self).__init__() + + final_oup = block_args.output_filters + + self._conv = ConvBNLayer(input_channels, + 1, + final_oup, + bn_act=None, + padding_type=padding_type, + name=name, + conv_name=name + "_project_conv", + bn_name="_bn2", + model_name=model_name, + cur_stage=cur_stage) + + def forward(self, inputs: paddle.Tensor): + return self._conv(inputs) + + +class SEBlock(nn.Layer): + """Basic Squeeze-and-Excitation block for Efficientnet.""" + def __init__(self, + input_channels: int, + num_squeezed_channels: int, + oup: int, + padding_type: str, + name: str = None, + model_name: str = None, + cur_stage: str = None): + super(SEBlock, self).__init__() + + self._pool = AdaptiveAvgPool2d(1) + self._conv1 = Conv2ds(input_channels, + num_squeezed_channels, + 1, + use_bias=True, + padding_type=padding_type, + act="swish", + name=name + "_se_reduce") + + self._conv2 = Conv2ds(num_squeezed_channels, + oup, + 1, + act="sigmoid", + use_bias=True, + padding_type=padding_type, + name=name + "_se_expand") + + def forward(self, inputs: paddle.Tensor): + x = self._pool(inputs) + x = self._conv1(x) + x = self._conv2(x) + return paddle.multiply(inputs, x) + + +class MbConvBlock(nn.Layer): + """Mobile inverted bottleneck convolution for Efficientnet.""" + def __init__(self, + input_channels: int, + block_args: dict, + padding_type: str, + use_se: bool, + name: str = None, + drop_connect_rate: float = None, + is_test: bool = False, + model_name: str = None, + cur_stage: str = None): + super(MbConvBlock, self).__init__() + + oup = block_args.input_filters * block_args.expand_ratio + self.block_args = block_args + self.has_se = use_se and (block_args.se_ratio is not None) and (0 < block_args.se_ratio <= 1) + self.id_skip = block_args.id_skip + self.expand_ratio = block_args.expand_ratio + self.drop_connect_rate = drop_connect_rate + self.is_test = is_test + + if self.expand_ratio != 1: + self._ecn = ExpandConvNorm(input_channels, + block_args, + padding_type=padding_type, + name=name, + model_name=model_name, + cur_stage=cur_stage) + + self._dcn = DepthwiseConvNorm(input_channels * block_args.expand_ratio, + block_args, + padding_type=padding_type, + name=name, + model_name=model_name, + cur_stage=cur_stage) + + if self.has_se: + num_squeezed_channels = max(1, int(block_args.input_filters * block_args.se_ratio)) + self._se = SEBlock(input_channels * block_args.expand_ratio, + num_squeezed_channels, + oup, + padding_type=padding_type, + name=name, + model_name=model_name, + cur_stage=cur_stage) + + self._pcn = ProjectConvNorm(input_channels * block_args.expand_ratio, + block_args, + padding_type=padding_type, + name=name, + model_name=model_name, + cur_stage=cur_stage) + + def forward(self, inputs: paddle.Tensor): + x = inputs + if self.expand_ratio != 1: + x = self._ecn(x) + x = F.swish(x) + x = self._dcn(x) + x = F.swish(x) + if self.has_se: + x = self._se(x) + x = self._pcn(x) + if self.id_skip and \ + self.block_args.stride == 1 and \ + self.block_args.input_filters == self.block_args.output_filters: + if self.drop_connect_rate: + x = _drop_connect(x, self.drop_connect_rate, self.is_test) + x = paddle.elementwise_add(x, inputs) + return x + + +class ConvStemNorm(nn.Layer): + """Basic conv stem norm block for extracting features.""" + def __init__(self, + input_channels: int, + padding_type: str, + _global_params: dict, + name: str = None, + model_name: str = None, + cur_stage: str = None): + super(ConvStemNorm, self).__init__() + + output_channels = round_filters(32, _global_params) + self._conv = ConvBNLayer(input_channels, + filter_size=3, + output_channels=output_channels, + stride=2, + bn_act=None, + padding_type=padding_type, + name="", + conv_name="_conv_stem", + bn_name="_bn0", + model_name=model_name, + cur_stage=cur_stage) + + def forward(self, inputs: paddle.Tensor): + return self._conv(inputs) + + +class ExtractFeatures(nn.Layer): + """Extract features.""" + def __init__(self, + input_channels: int, + _block_args: dict, + _global_params: dict, + padding_type: str, + use_se: bool, + is_test: bool, + model_name: str = None): + super(ExtractFeatures, self).__init__() + + self._global_params = _global_params + + self._conv_stem = ConvStemNorm(input_channels, + padding_type=padding_type, + _global_params=_global_params, + model_name=model_name, + cur_stage=0) + + self.block_args_copy = copy.deepcopy(_block_args) + idx = 0 + block_size = 0 + for block_arg in self.block_args_copy: + block_arg = block_arg._replace(input_filters=round_filters(block_arg.input_filters, _global_params), + output_filters=round_filters(block_arg.output_filters, _global_params), + num_repeat=round_repeats(block_arg.num_repeat, _global_params)) + block_size += 1 + for _ in range(block_arg.num_repeat - 1): + block_size += 1 + + self.conv_seq = [] + cur_stage = 1 + for block_args in _block_args: + block_args = block_args._replace(input_filters=round_filters(block_args.input_filters, _global_params), + output_filters=round_filters(block_args.output_filters, _global_params), + num_repeat=round_repeats(block_args.num_repeat, _global_params)) + + drop_connect_rate = self._global_params.drop_connect_rate if not is_test else 0 + if drop_connect_rate: + drop_connect_rate *= float(idx) / block_size + + _mc_block = self.add_sublayer( + "_blocks." + str(idx) + ".", + MbConvBlock(block_args.input_filters, + block_args=block_args, + padding_type=padding_type, + use_se=use_se, + name="_blocks." + str(idx) + ".", + drop_connect_rate=drop_connect_rate, + model_name=model_name, + cur_stage=cur_stage)) + self.conv_seq.append(_mc_block) + idx += 1 + if block_args.num_repeat > 1: + block_args = block_args._replace(input_filters=block_args.output_filters, stride=1) + for _ in range(block_args.num_repeat - 1): + drop_connect_rate = self._global_params.drop_connect_rate if not is_test else 0 + if drop_connect_rate: + drop_connect_rate *= float(idx) / block_size + _mc_block = self.add_sublayer( + "block." + str(idx) + ".", + MbConvBlock(block_args.input_filters, + block_args, + padding_type=padding_type, + use_se=use_se, + name="_blocks." + str(idx) + ".", + drop_connect_rate=drop_connect_rate, + model_name=model_name, + cur_stage=cur_stage)) + self.conv_seq.append(_mc_block) + idx += 1 + cur_stage += 1 + + def forward(self, inputs: paddle.Tensor): + x = self._conv_stem(inputs) + x = F.swish(x) + for _mc_block in self.conv_seq: + x = _mc_block(x) + return x + + +@moduleinfo(name="efficientnetb7_imagenet", + type="cv/classification", + author="paddlepaddle", + author_email="", + summary="efficientnetb7_imagenet is a classification model, " + "this module is trained with Imagenet dataset.", + version="1.1.0", + meta=ImageClassifierModule) +class EfficientNet_B7(nn.Layer): + def __init__(self, + is_test: bool = False, + padding_type: str = "SAME", + override_params: dict = None, + use_se: bool = True, + class_dim: int = 1000, + load_checkpoint: str = None): + super(EfficientNet_B7, self).__init__() + + model_name = 'efficientnet-b7' + self.name = "b7" + self._block_args, self._global_params = get_model_params(model_name, override_params) + self.padding_type = padding_type + self.use_se = use_se + self.is_test = is_test + + self._ef = ExtractFeatures(3, + self._block_args, + self._global_params, + self.padding_type, + self.use_se, + self.is_test, + model_name=self.name) + + output_channels = round_filters(1280, self._global_params) + oup = 640 + + self._conv = ConvBNLayer(oup, + 1, + output_channels, + bn_act="swish", + padding_type=self.padding_type, + name="", + conv_name="_conv_head", + bn_name="_bn1", + model_name=self.name, + cur_stage=7) + self._pool = AdaptiveAvgPool2d(1) + + if self._global_params.dropout_rate: + self._drop = Dropout(p=self._global_params.dropout_rate, mode="upscale_in_train") + + param_attr, bias_attr = init_fc_layer("_fc") + self._fc = Linear(output_channels, class_dim, weight_attr=param_attr, bias_attr=bias_attr) + + if load_checkpoint is not None: + model_dict = paddle.load(load_checkpoint)[0] + self.set_dict(model_dict) + print("load custom checkpoint success") + + else: + checkpoint = os.path.join(self.directory, 'efficientnet_b7_imagenet.pdparams') + if not os.path.exists(checkpoint): + os.system( + 'wget https://paddlehub.bj.bcebos.com/dygraph/image_classification/efficientnet_b7_imagenet.pdparams -O ' + + checkpoint) + model_dict = paddle.load(checkpoint)[0] + self.set_dict(model_dict) + print("load pretrained checkpoint success") + + def forward(self, inputs: paddle.Tensor): + x = self._ef(inputs) + x = self._conv(x) + x = self._pool(x) + if self._global_params.dropout_rate: + x = self._drop(x) + x = paddle.squeeze(x, axis=[2, 3]) + x = self._fc(x) + return x diff --git a/hub_module/modules/image/classification/efficientnetb7_imagenet/processor.py b/hub_module/modules/image/classification/efficientnetb7_imagenet/processor.py deleted file mode 100644 index 12ccc9b2..00000000 --- a/hub_module/modules/image/classification/efficientnetb7_imagenet/processor.py +++ /dev/null @@ -1,69 +0,0 @@ -# -*- coding:utf-8 -*- -# copyright (c) 2020 PaddlePaddle Authors. All Rights Reserve. -# -# Licensed under the Apache License, Version 2.0 (the "License"); -# you may not use this file except in compliance with the License. -# You may obtain a copy of the License at -# -# http://www.apache.org/licenses/LICENSE-2.0 -# -# Unless required by applicable law or agreed to in writing, software -# distributed under the License is distributed on an "AS IS" BASIS, -# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. -# See the License for the specific language governing permissions and -# limitations under the License. - -from __future__ import absolute_import -from __future__ import division -from __future__ import print_function - -import base64 -import cv2 -import os - -import numpy as np - - -def base64_to_cv2(b64str): - data = base64.b64decode(b64str.encode('utf8')) - data = np.fromstring(data, np.uint8) - data = cv2.imdecode(data, cv2.IMREAD_COLOR) - return data - - -def softmax(x): - orig_shape = x.shape - if len(x.shape) > 1: - tmp = np.max(x, axis=1) - x -= tmp.reshape((x.shape[0], 1)) - x = np.exp(x) - tmp = np.sum(x, axis=1) - x /= tmp.reshape((x.shape[0], 1)) - else: - tmp = np.max(x) - x -= tmp - x = np.exp(x) - tmp = np.sum(x) - x /= tmp - return x - - -def postprocess(data_out, label_list, top_k): - """ - Postprocess output of network, one image at a time. - - Args: - data_out (numpy.ndarray): output data of network. - label_list (list): list of label. - top_k (int): Return top k results. - """ - output = [] - for result in data_out: - result_i = softmax(result) - output_i = {} - indexs = np.argsort(result_i)[::-1][0:top_k] - for index in indexs: - label = label_list[index].split(',')[0] - output_i[label] = float(result_i[index]) - output.append(output_i) - return output -- GitLab