Update gan1.6 (#3551)

* replace instance_norm

* add version check

* update check version

* -1->None

* fix mistake

PaddleCV/PaddleGAN/README.md

@@ -21,7 +21,7 @@
 注意：
 1. StarGAN，AttGAN和STGAN由于梯度惩罚所需的操作目前只支持GPU，需使用GPU训练。
 2. GAN模型目前仅仅验证了单机单卡训练和预测结果。
-3. CGAN和DCGAN两个模型训练使用的数据集为MNIST数据集；StarGAN，AttGAN和STGAN的数据集为CelebA数据集。Pix2Pix和CycleGAN支持的数据集可以参考download.py中的cycle_pix_dataset。
+3. CGAN和DCGAN两个模型训练使用的数据集为MNIST数据集；StarGAN，AttGAN和STGAN的数据集为CelebA数据集。Pix2Pix和CycleGAN支持的数据集可以参考download.py中的cycle_pix_dataset。cityscapes数据集需要从[官方](https://www.cityscapes-dataset.com)下载数据，下载完之后使用`scripts/prepare_cityscapes_dataset.py`处理，处理后的文件夹命名为cityscapes并放入data目录下即可。
 4. PaddlePaddle1.5.1及之前的版本不支持在AttGAN和STGAN模型里的判别器加上的instance norm。如果要在判别器中加上instance norm，请源码编译develop分支并安装。
 5. 中间效果图保存在${output_dir}/test文件夹中。对于Pix2Pix来说，inputA 和inputB 代表输入的两种风格的图片，fakeB表示生成图片；对于CycleGAN来说，inputA表示输入图片，fakeB表示inputA根据生成的图片，cycA表示fakeB经过生成器重构出来的对应于inputA的重构图片；对于StarGAN，AttGAN和STGAN来说，第一行表示原图，之后的每一行都代表一种属性变换。
 6. infer过程使用的test_list文件和训练过程中使用的train_list具有相同格式，第一行为样本数量，第二行为属性，之后的行中第一个表示图片名称，之后的-1和1表示该图片是否拥有该属性(1为有该属性，-1为没有该属性)。
@@ -67,7 +67,7 @@
 ### 安装说明
 **安装[PaddlePaddle](https://github.com/PaddlePaddle/Paddle)：**
 
-在当前目录下运行样例代码需要PadddlePaddle Fluid的v.1.5或以上的版本。如果你的运行环境中的PaddlePaddle低于此版本，请根据[安装文档](https://www.paddlepaddle.org.cn/documentation/docs/zh/1.5/beginners_guide/install/index_cn.html)中的说明来更新PaddlePaddle。
+在当前目录下运行样例代码需要PadddlePaddle Fluid的v.1.6或以上的版本。如果你的运行环境中的PaddlePaddle低于此版本，请根据[安装文档](https://www.paddlepaddle.org.cn/documentation/docs/zh/1.5/beginners_guide/install/index_cn.html)中的说明来更新PaddlePaddle。
 
 其他依赖包：
 1. `pip install imageio` 或者 `pip install -r requirements.txt` 安装imageio包（保存图片代码中所依赖的包）

PaddleCV/PaddleGAN/download.py

@@ -153,8 +153,8 @@ if __name__ == '__main__':
     args = parser.parse_args()
     cycle_pix_dataset = [
         'apple2orange', 'summer2winter_yosemite', 'horse2zebra', 'monet2photo',
-        'cezanne2photo', 'ukiyoe2photo', 'vangogh2photo', 'maps', 'cityscapes',
-        'facades', 'iphone2dslr_flower', 'ae_photos', 'mini'
+        'cezanne2photo', 'ukiyoe2photo', 'vangogh2photo', 'maps', 'facades',
+        'iphone2dslr_flower', 'ae_photos', 'mini'
     ]
 
     pwd = os.path.join(os.path.dirname(__file__), 'data')

PaddleCV/PaddleGAN/infer.py

@@ -27,7 +27,7 @@ import imageio
 import glob
 from util.config import add_arguments, print_arguments
 from data_reader import celeba_reader_creator, reader_creator, triplex_reader_creator
-from util.utility import check_attribute_conflict, check_gpu, save_batch_image
+from util.utility import check_attribute_conflict, check_gpu, save_batch_image, check_version
 from util import utility
 import copy
 
@@ -82,6 +82,7 @@ def infer(args):
         name='image_name', shape=[args.n_samples], dtype='int32')
 
     model_name = 'net_G'
+
     if args.model_net == 'CycleGAN':
         loader = fluid.io.DataLoader.from_generator(
             feed_list=[input, image_name],
@@ -383,4 +384,5 @@ if __name__ == "__main__":
     args = parser.parse_args()
     print_arguments(args)
     check_gpu(args.use_gpu)
+    check_version()
     infer(args)

PaddleCV/PaddleGAN/network/base_network.py

@@ -64,12 +64,6 @@ def norm_layer(input,
             moving_variance_name=name + '_var')
 
     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"
@@ -91,15 +85,8 @@ def norm_layer(input,
                 name=offset_name,
                 initializer=fluid.initializer.Constant(0.0),
                 trainable=False)
-        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
+        return fluid.layers.instance_norm(
+            input, param_attr=scale_param, bias_attr=offset_param)
     else:
         raise NotImplementedError("norm type: [%s] is not support" % norm_type)
 

PaddleCV/PaddleGAN/scripts/prepare_cityscapes_dataset.py

+import os
+import argparse
+import functools
+import glob
+from PIL import Image
+''' Based on https://github.com/junyanz/CycleGAN'''
+
+
+def load_image(path):
+    return Image.open(path).convert('RGB').resize((256, 256))
+
+
+def propress_cityscapes(gtFine_dir, leftImg8bit_dir, output_dir, phase):
+    save_dir = os.path.join(output_dir, phase)
+    try:
+        os.makedirs(save_dir)
+    except Exception as e:
+        print("{} makedirs".format(e))
+        pass
+    try:
+        os.makedirs(os.path.join(save_dir, 'A'))
+    except Exception as e:
+        print("{} makedirs".format(e))
+    try:
+        os.makedirs(os.path.join(save_dir, 'B'))
+    except Exception as e:
+        print("{} makedirs".format(e))
+
+    seg_expr = os.path.join(gtFine_dir, phase, "*", "*_color.png")
+    seg_paths = glob.glob(seg_expr)
+    seg_paths = sorted(seg_paths)
+
+    photo_expr = os.path.join(leftImg8bit_dir, phase, "*", '*_leftImg8bit.png')
+    photo_paths = glob.glob(photo_expr)
+    photo_paths = sorted(photo_paths)
+
+    assert len(seg_paths) == len(photo_paths), \
+          "[%d] gtFine images NOT match [%d] leftImg8bit images. Aborting." % (len(segmap_paths), len(photo_paths))
+
+    for i, (seg_path, photo_path) in enumerate(zip(seg_paths, photo_paths)):
+        seg_image = load_image(seg_path)
+        photo_image = load_image(photo_path)
+        # save image
+        save_path = os.path.join(save_dir, 'A', "%d_A.jpg" % i)
+        photo_image.save(save_path, format='JPEG', subsampling=0, quality=100)
+        save_path = os.path.join(save_dir, 'B', "%d_B.jpg" % i)
+        seg_image.save(save_path, format='JPEG', subsampling=0, quality=100)
+
+        if i % 10 == 0:
+            print("preprocess %d ~ %d images." % (i, i + 10))
+
+
+if __name__ == '__main__':
+    parser = argparse.ArgumentParser(description=__doc__)
+    # yapf: disable
+    parser.add_argument('--gtFine_dir',       type=str,     default=None,       help='Path to Cityscapes gtFine directory.')
+    parser.add_argument('--leftImg8bit_dir',  type=str,     default=None,       help='Path to Cityscapes leftImg8bit_trainvaltest directory.')
+    parser.add_argument('--output_dir',       type=str,     default=None,       help='Path to output Cityscapes directory.')
+    # yapf: enable
+    args = parser.parse_args()
+
+    print('Preparing Cityscapes Dataset for val phase')
+    propress_cityscapes(args.gtFine_dir, args.leftImg8bit_dir, args.output_dir,
+                        'val')
+
+    print('Preparing Cityscapes Dataset for train phase')
+    propress_cityscapes(args.gtFine_dir, args.leftImg8bit_dir, args.output_dir,
+                        'train')
+
+    print("DONE!!!")

PaddleCV/PaddleGAN/scripts/run_attgan.sh

-python train.py --model_net AttGAN --dataset celeba --crop_size 170 --image_size 128 --train_list ./data/celeba/list_attr_celeba.txt  --gan_mode wgan  --batch_size 32 --print_freq 1 --num_discriminator_time 5 --epoch 120 --output ./output/attgan/ >log_out 2>log_err
+python train.py --model_net AttGAN --dataset celeba --crop_size 170 --image_size 128 --train_list ./data/celeba/list_attr_celeba.txt  --gan_mode wgan  --batch_size 32 --print_freq 1 --num_discriminator_time 5 --epoch 120 --dis_norm instance_norm --output ./output/attgan/ >log_out 2>log_err

PaddleCV/PaddleGAN/scripts/run_stgan.sh

-python train.py --model_net STGAN --dataset celeba --crop_size 170 --image_size 128 --train_list ./data/celeba/list_attr_celeba.txt --gan_mode wgan  --batch_size 32 --print_freq 1 --num_discriminator_time 5 --epoch 50 --output ./output/stgan/ >log_out 2>log_err
+python train.py --model_net STGAN --dataset celeba --crop_size 170 --image_size 128 --train_list ./data/celeba/list_attr_celeba.txt --gan_mode wgan  --batch_size 32 --print_freq 1 --num_discriminator_time 5 --epoch 50 --dis_norm instance_norm --output ./output/stgan/ >log_out 2>log_err

PaddleCV/PaddleGAN/train.py

@@ -30,7 +30,8 @@ import trainer
 def train(cfg):
 
     MODELS = [
-        "CGAN", "DCGAN", "Pix2pix", "CycleGAN", "StarGAN", "AttGAN", "STGAN", "SPADE"
+        "CGAN", "DCGAN", "Pix2pix", "CycleGAN", "StarGAN", "AttGAN", "STGAN",
+        "SPADE"
     ]
     if cfg.model_net not in MODELS:
         raise NotImplementedError("{} is not support!".format(cfg.model_net))
@@ -65,6 +66,7 @@ if __name__ == "__main__":
     cfg = config.parse_args()
     config.print_arguments(cfg)
     utility.check_gpu(cfg.use_gpu)
+    utility.check_version()
     if cfg.profile:
         if cfg.use_gpu:
             with fluid.profiler.profiler('All', 'total',

PaddleCV/PaddleGAN/trainer/AttGAN.py

@@ -270,18 +270,18 @@ class AttGAN(object):
         self.id2name = id2name
 
     def build_model(self):
-        data_shape = [-1, 3, self.cfg.image_size, self.cfg.image_size]
+        data_shape = [None, 3, self.cfg.image_size, self.cfg.image_size]
 
         image_real = fluid.data(
             name='image_real', shape=data_shape, dtype='float32')
         label_org = fluid.data(
-            name='label_org', shape=[-1, self.cfg.c_dim], dtype='float32')
+            name='label_org', shape=[None, self.cfg.c_dim], dtype='float32')
         label_trg = fluid.data(
-            name='label_trg', shape=[-1, self.cfg.c_dim], dtype='float32')
+            name='label_trg', shape=[None, self.cfg.c_dim], dtype='float32')
         label_org_ = fluid.data(
-            name='label_org_', shape=[-1, self.cfg.c_dim], dtype='float32')
+            name='label_org_', shape=[None, self.cfg.c_dim], dtype='float32')
         label_trg_ = fluid.data(
-            name='label_trg_', shape=[-1, self.cfg.c_dim], dtype='float32')
+            name='label_trg_', shape=[None, self.cfg.c_dim], dtype='float32')
 
         py_reader = fluid.io.PyReader(
             feed_list=[image_real, label_org, label_trg],
@@ -369,9 +369,9 @@ class AttGAN(object):
                 batch_id += 1
 
             if self.cfg.run_test:
-                image_name = fluid.layers.data(
+                image_name = fluid.data(
                     name='image_name',
-                    shape=[self.cfg.n_samples],
+                    shape=[None, self.cfg.n_samples],
                     dtype='int32')
                 test_py_reader = fluid.io.PyReader(
                     feed_list=[image_real, label_org, label_trg, image_name],

PaddleCV/PaddleGAN/trainer/CGAN.py

@@ -88,11 +88,12 @@ class CGAN(object):
 
     def build_model(self):
 
-        img = fluid.data(name='img', shape=[-1, 784], dtype='float32')
-        condition = fluid.data(name='condition', shape=[-1, 1], dtype='float32')
+        img = fluid.data(name='img', shape=[None, 784], dtype='float32')
+        condition = fluid.data(
+            name='condition', shape=[None, 1], dtype='float32')
         noise = fluid.data(
-            name='noise', shape=[-1, self.cfg.noise_size], dtype='float32')
-        label = fluid.data(name='label', shape=[-1, 1], dtype='float32')
+            name='noise', shape=[None, self.cfg.noise_size], dtype='float32')
+        label = fluid.data(name='label', shape=[None, 1], dtype='float32')
 
         g_trainer = GTrainer(noise, condition, self.cfg)
         d_trainer = DTrainer(img, condition, label, self.cfg)

PaddleCV/PaddleGAN/trainer/CycleGAN.py

@@ -229,7 +229,7 @@ class CycleGAN(object):
         self.B_id2name = B_id2name
 
     def build_model(self):
-        data_shape = [-1, 3, self.cfg.crop_size, self.cfg.crop_size]
+        data_shape = [None, 3, self.cfg.crop_size, self.cfg.crop_size]
 
         input_A = fluid.data(name='input_A', shape=data_shape, dtype='float32')
         input_B = fluid.data(name='input_B', shape=data_shape, dtype='float32')
@@ -347,9 +347,9 @@ class CycleGAN(object):
 
             if self.cfg.run_test:
                 A_image_name = fluid.data(
-                    name='A_image_name', shape=[-1, 1], dtype='int32')
+                    name='A_image_name', shape=[None, 1], dtype='int32')
                 B_image_name = fluid.data(
-                    name='B_image_name', shape=[-1, 1], dtype='int32')
+                    name='B_image_name', shape=[None, 1], dtype='int32')
                 A_test_py_reader = fluid.io.PyReader(
                     feed_list=[input_A, A_image_name],
                     capacity=4,

PaddleCV/PaddleGAN/trainer/DCGAN.py

@@ -86,10 +86,10 @@ class DCGAN(object):
         self.train_reader = train_reader
 
     def build_model(self):
-        img = fluid.data(name='img', shape=[-1, 784], dtype='float32')
+        img = fluid.data(name='img', shape=[None, 784], dtype='float32')
         noise = fluid.data(
-            name='noise', shape=[-1, self.cfg.noise_size], dtype='float32')
-        label = fluid.data(name='label', shape=[-1, 1], dtype='float32')
+            name='noise', shape=[None, self.cfg.noise_size], dtype='float32')
+        label = fluid.data(name='label', shape=[None, 1], dtype='float32')
 
         g_trainer = GTrainer(noise, label, self.cfg)
         d_trainer = DTrainer(img, label, self.cfg)

PaddleCV/PaddleGAN/trainer/Pix2pix.py

@@ -211,7 +211,7 @@ class Pix2pix(object):
         self.id2name = id2name
 
     def build_model(self):
-        data_shape = [-1, 3, self.cfg.crop_size, self.cfg.crop_size]
+        data_shape = [None, 3, self.cfg.crop_size, self.cfg.crop_size]
 
         input_A = fluid.data(name='input_A', shape=data_shape, dtype='float32')
         input_B = fluid.data(name='input_B', shape=data_shape, dtype='float32')
@@ -298,7 +298,7 @@ class Pix2pix(object):
             if self.cfg.run_test:
                 image_name = fluid.data(
                     name='image_name',
-                    shape=[-1, self.cfg.batch_size],
+                    shape=[None, self.cfg.batch_size],
                     dtype="int32")
                 test_loader = fluid.io.DataLoader.from_generator(
                     feed_list=[input_A, input_B, image_name],

PaddleCV/PaddleGAN/trainer/SPADE.py

@@ -283,11 +283,11 @@ class SPADE(object):
         self.id2name = id2name
 
     def build_model(self):
-        data_shape = [-1, 3, self.cfg.crop_height, self.cfg.crop_width]
+        data_shape = [None, 3, self.cfg.crop_height, self.cfg.crop_width]
         label_shape = [
-            -1, self.cfg.label_nc, self.cfg.crop_height, self.cfg.crop_width
+            None, self.cfg.label_nc, self.cfg.crop_height, self.cfg.crop_width
         ]
-        edge_shape = [-1, 1, self.cfg.crop_height, self.cfg.crop_width]
+        edge_shape = [None, 1, self.cfg.crop_height, self.cfg.crop_width]
 
         input_A = fluid.data(
             name='input_label', shape=label_shape, dtype='float32')
@@ -389,7 +389,7 @@ class SPADE(object):
                 test_program = gen_trainer.infer_program
                 image_name = fluid.data(
                     name='image_name',
-                    shape=[-1, self.cfg.batch_size],
+                    shape=[None, self.cfg.batch_size],
                     dtype="int32")
                 test_py_reader = fluid.io.PyReader(
                     feed_list=[input_A, input_B, input_C, image_name],

PaddleCV/PaddleGAN/trainer/STGAN.py

@@ -282,18 +282,18 @@ class STGAN(object):
         self.batch_num = batch_num
 
     def build_model(self):
-        data_shape = [-1, 3, self.cfg.image_size, self.cfg.image_size]
+        data_shape = [None, 3, self.cfg.image_size, self.cfg.image_size]
 
         image_real = fluid.data(
             name='image_real', shape=data_shape, dtype='float32')
         label_org = fluid.data(
-            name='label_org', shape=[-1, self.cfg.c_dim], dtype='float32')
+            name='label_org', shape=[None, self.cfg.c_dim], dtype='float32')
         label_trg = fluid.data(
-            name='label_trg', shape=[-1, self.cfg.c_dim], dtype='float32')
+            name='label_trg', shape=[None, self.cfg.c_dim], dtype='float32')
         label_org_ = fluid.data(
-            name='label_org_', shape=[-1, self.cfg.c_dim], dtype='float32')
+            name='label_org_', shape=[None, self.cfg.c_dim], dtype='float32')
         label_trg_ = fluid.data(
-            name='label_trg_', shape=[-1, self.cfg.c_dim], dtype='float32')
+            name='label_trg_', shape=[None, self.cfg.c_dim], dtype='float32')
 
         test_gen_trainer = GTrainer(image_real, label_org, label_org_,
                                     label_trg, label_trg_, self.cfg,
@@ -380,7 +380,7 @@ class STGAN(object):
             if self.cfg.run_test:
                 image_name = fluid.data(
                     name='image_name',
-                    shape=[-1, self.cfg.n_samples],
+                    shape=[None, self.cfg.n_samples],
                     dtype='int32')
                 test_py_reader = fluid.io.PyReader(
                     feed_list=[image_real, label_org, label_trg, image_name],

PaddleCV/PaddleGAN/trainer/StarGAN.py

@@ -259,14 +259,14 @@ class StarGAN(object):
         self.batch_num = batch_num
 
     def build_model(self):
-        data_shape = [-1, 3, self.cfg.image_size, self.cfg.image_size]
+        data_shape = [None, 3, self.cfg.image_size, self.cfg.image_size]
 
         image_real = fluid.data(
             name='image_real', shape=data_shape, dtype='float32')
         label_org = fluid.data(
-            name='label_org', shape=[-1, self.cfg.c_dim], dtype='float32')
+            name='label_org', shape=[None, self.cfg.c_dim], dtype='float32')
         label_trg = fluid.data(
-            name='label_trg', shape=[-1, self.cfg.c_dim], dtype='float32')
+            name='label_trg', shape=[None, self.cfg.c_dim], dtype='float32')
 
         py_reader = fluid.io.PyReader(
             feed_list=[image_real, label_org, label_trg],
@@ -348,7 +348,7 @@ class StarGAN(object):
             if self.cfg.run_test:
                 image_name = fluid.data(
                     name='image_name',
-                    shape=[-1, self.cfg.n_samples],
+                    shape=[None, self.cfg.n_samples],
                     dtype='int32')
                 test_py_reader = fluid.io.PyReader(
                     feed_list=[image_real, label_org, label_trg, image_name],

PaddleCV/PaddleGAN/util/utility.py

@@ -409,3 +409,19 @@ def check_gpu(use_gpu):
             sys.exit(1)
     except Exception as e:
         pass
+
+
+def check_version():
+    """
+    Log error and exit when the installed version of paddlepaddle is
+    not satisfied.
+    """
+    err = "PaddlePaddle version 1.6 or higher is required, " \
+          "or a suitable develop version is satisfied as well. \n" \
+          "Please make sure the version is good with your code." \
+
+    try:
+        fluid.require_version('1.6.0')
+    except Exception as e:
+        print(err)
+        sys.exit(1)