Merge pull request #448 from wangjiawei04/pddet

Add paddle detection support

core/general-client/src/pybind_general_model.cpp

@@ -78,7 +78,6 @@ PYBIND11_MODULE(serving_client, m) {
            [](PredictorClient &self) { self.create_predictor(); })
       .def("destroy_predictor",
            [](PredictorClient &self) { self.destroy_predictor(); })
-
       .def("batch_predict",
            [](PredictorClient &self,
               const std::vector<std::vector<std::vector<float>>>

python/examples/faster_rcnn_model/README.md

+# Faster RCNN model on Paddle Serving
+
+([简体中文](./README_CN.md)|English)
+
+This article requires [Paddle Detection](https://github.com/PaddlePaddle/PaddleDetection) trained models and configuration files. If users want to quickly deploy on Paddle Serving, please read the Chapter 2 directly.
+
+## 1. Train an object detection model
+
+Users can read [Paddle Detection Getting Started](https://github.com/PaddlePaddle/PaddleDetection/blob/release/0.2/docs/tutorials/GETTING_STARTED_cn.md) to understand the background of Paddle Detection. The purpose of PaddleDetection is to provide a rich and easy-to-use object detection model for industry and academia. Not only is it superior in performance and easy to deploy, but it can also flexibly meet the needs of algorithm research.
+
+### Environmental requirements
+
+CPU version: No special requirements
+
+GPU version: CUDA 9.0 and above
+
+```
+git clone https://github.com/PaddlePaddle/PaddleDetection
+cd PaddleDetection
+```
+Next, you can train the faster rcnn model
+```
+python tools/train.py -c configs/faster_rcnn_r50_1x.yml
+```
+The time for training the model depends on the situation and is related to the computing power of the training equipment and the number of iterations.
+In the training process, `faster_rcnn_r50_1x.yml` defines the snapshot of the saved model. After the final training, the model with the best effect will be saved as `best_model.pdmodel`, which is a compressed PaddleDetection Exclusive model files.
+
+**If we want the model to be used by Paddle Serving, we must do export_model.**
+
+Output model
+```
+python export_model.py
+```
+## 2. Start the model and predict
+If users do not use the Paddle Detection project to train models, we are here to provide you with sample model downloads. If you trained the model with Paddle Detection, you can skip the ** Download Model ** section.
+
+### Download model
+```
+wget https://paddle-serving.bj.bcebos.com/pddet_demo/faster_rcnn_model.tar.gz
+wget https://paddle-serving.bj.bcebos.com/pddet_demo/paddle_serving_app-0.0.1-py2-none-any.whl
+wget https://paddle-serving.bj.bcebos.com/pddet_demo/infer_cfg.yml
+tar xf faster_rcnn_model.tar.gz
+mv faster_rcnn_model/pddet *.
+```
+
+### Start the service
+```
+GLOG_v = 2 python -m paddle_serving_server_gpu.serve --model pddet_serving_model --port 9494 --gpu_id 0
+```
+
+### Perform prediction
+```
+python test_client.py --config_path = infer_cfg.yml --infer_img = 000000570688.jpg --dump_result --visualize
+```
+
+## 3. Result analysis
+<p align = "center">
+    <br>
+<img src = '000000570688.jpg'>
+    <br>
+<p>
+This is the input picture
+  
+<p align = "center">
+    <br>
+<img src = '000000570688_bbox.jpg'>
+    <br>
+<p>
+  
+This is the picture after adding bbox. You can see that the client has done post-processing for the picture. In addition, the output/bbox.json also has the number and coordinate information of each box.

python/examples/faster_rcnn_model/README_CN.md

+# Faster RCNN模型
+
+(简体中文|[English](./README.md))
+
+本文需要[Paddle Detection](https://github.com/PaddlePaddle/PaddleDetection)训练的模型和配置文件。如果用户想要快速部署在Paddle Serving上，请直接阅读第二章节。
+
+## 1. 训练物体检测模型
+
+用户可以阅读 [Paddle Detection入门使用](https://github.com/PaddlePaddle/PaddleDetection/blob/release/0.2/docs/tutorials/GETTING_STARTED_cn.md)来了解Paddle Detection的背景。PaddleDetection的目的是为工业界和学术界提供丰富、易用的目标检测模型。不仅性能优越、易于部署，而且能够灵活的满足算法研究的需求。
+
+### 环境要求
+
+CPU版： 没有特别要求
+
+GPU版： CUDA 9.0及以上
+
+```
+git clone https://github.com/PaddlePaddle/PaddleDetection
+cd PaddleDetection
+```
+接下来可以训练faster rcnn模型
+```
+python tools/train.py -c configs/faster_rcnn_r50_1x.yml
+```
+训练模型的时间视情况而定，与训练的设备算力和迭代轮数相关。
+在训练的过程中，`faster_rcnn_r50_1x.yml`当中定义了保存模型的`snapshot`，在最终训练完成后，效果最好的模型，会被保存为`best_model.pdmodel`，这是一个经过压缩的PaddleDetection的专属模型文件。
+
+**如果我们要让模型可被Paddle Serving所使用，必须做export_model。**
+
+输出模型
+```
+python export_model.py
+```
+
+## 2. 启动模型并预测
+如果用户没有用Paddle Detection项目训练模型，我们也在此为您提供示例模型下载。如果您用Paddle Detection训练了模型，可以跳过 **下载模型** 部分。
+
+### 下载模型
+```
+wget https://paddle-serving.bj.bcebos.com/pddet_demo/faster_rcnn_model.tar.gz
+wget https://paddle-serving.bj.bcebos.com/pddet_demo/paddle_serving_app-0.0.1-py2-none-any.whl
+wget https://paddle-serving.bj.bcebos.com/pddet_demo/infer_cfg.yml
+tar xf faster_rcnn_model.tar.gz
+mv faster_rcnn_model/pddet* .
+```
+
+### 启动服务
+```
+GLOG_v=2 python -m paddle_serving_server_gpu.serve --model pddet_serving_model --port 9494 --gpu_id 0
+```
+
+### 执行预测
+```
+python test_client.py --config_path=infer_cfg.yml --infer_img=000000570688.jpg --dump_result --visualize
+```
+
+## 3. 结果分析
+<p align="center">
+    <br>
+<img src='000000570688.jpg' >
+    <br>
+<p>
+这是输入图片
+  
+<p align="center">
+    <br>
+<img src='000000570688_bbox.jpg' >
+    <br>
+<p>
+这是实现添加了bbox之后的图片，可以看到客户端已经为图片做好了后处理，此外在output/bbox.json也有各个框的编号和坐标信息。

python/examples/faster_rcnn_model/test_client.py

+# Copyright (c) 2020 PaddlePaddle Authors. All Rights Reserved.
+#
+# 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 paddle_serving_client import Client
+import sys
+import os
+import time
+from paddle_serving_app.reader.pddet import Detection
+import numpy as np
+
+py_version = sys.version_info[0]
+
+feed_var_names = ['image', 'im_shape', 'im_info']
+fetch_var_names = ['multiclass_nms']
+pddet = Detection(config_path=sys.argv[2], output_dir="./output")
+feed_dict = pddet.preprocess(feed_var_names, sys.argv[3])
+client = Client()
+client.load_client_config(sys.argv[1])
+client.connect(['127.0.0.1:9494'])
+fetch_map = client.predict(feed=feed_dict, fetch=fetch_var_names)
+outs = fetch_map.values()
+pddet.postprocess(fetch_map, fetch_var_names)

python/paddle_serving_app/reader/pddet/__init__.py

+# Copyright (c) 2020 PaddlePaddle Authors. All Rights Reserved.
+#
+# 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
+import argparse
+from .image_tool import Resize, Detection

python/setup.py.app.in

@@ -47,7 +47,8 @@ REQUIRED_PACKAGES = [
 
 packages=['paddle_serving_app',
           'paddle_serving_app.reader',
-	  'paddle_serving_app.utils']
+	  'paddle_serving_app.utils',
+      'paddle_serving_app.reader.pddet']
 
 package_data={}
 package_dir={'paddle_serving_app':
@@ -55,7 +56,9 @@ package_dir={'paddle_serving_app':
              'paddle_serving_app.reader':
              '${PADDLE_SERVING_BINARY_DIR}/python/paddle_serving_app/reader',
 	     'paddle_serving_app.utils':
-	     '${PADDLE_SERVING_BINARY_DIR}/python/paddle_serving_app/utils',}
+	     '${PADDLE_SERVING_BINARY_DIR}/python/paddle_serving_app/utils',
+          'paddle_serving_app.reader.pddet':
+           '${PADDLE_SERVING_BINARY_DIR}/python/paddle_serving_app/reader/pddet',}
 
 setup(
     name='paddle-serving-app',