基于vgg网路和TensorFlow的图像风格迁移系统源码+文档说明.zip

# tf2-neural-style-transfer 使用Tensorflow 2.0实现的图片风格迁移Demo。 给定一张内容图片和一张风格图片，生成一张具备内容图片内容+风格图片风格的图片。 风格图片示例：  示例1： 输入图片  输出图片  示例2： 输入图片  输出图片  # 快速开始 开发环境: Python 3.7 (Anaconda安装的) 1.clone 项目 2.切换到子项目根目录： ```cd DeepLearningExamples/tf2-neural-style-transfer``` 3.安装依赖环境： ```pip install -r requirements.txt``` 4.根据个人需求，修改`settings.py`中的配置参数。 5.运行训练脚本： ```python train.py``` 运行成功则会输出训练进度： ``` Epoch 1/20: 100%|██████████| 100/100 [00:09<00:00, 10.37it/s, loss=22066.4688] Epoch 2/20: 100%|██████████| 100/100 [00:07<00:00, 13.09it/s, loss=12479.9658] Epoch 3/20: 100%|██████████| 100/100 [00:07<00:00, 13.06it/s, loss=9065.9258] Epoch 4/20: 100%|██████████| 100/100 [00:07<00:00, 13.06it/s, loss=6993.3652] Epoch 5/20: 100%|██████████| 100/100 [00:07<00:00, 13.02it/s, loss=5558.0947] Epoch 6/20: 100%|██████████| 100/100 [00:07<00:00, 13.03it/s, loss=4526.5439] Epoch 7/20: 100%|██████████| 100/100 [00:07<00:00, 13.01it/s, loss=3777.7947] Epoch 8/20: 100%|██████████| 100/100 [00:07<00:00, 13.00it/s, loss=3228.8064] Epoch 9/20: 100%|██████████| 100/100 [00:07<00:00, 12.98it/s, loss=2821.5425] Epoch 10/20: 100%|██████████| 100/100 [00:07<00:00, 12.98it/s, loss=2515.5208] Epoch 11/20: 100%|██████████| 100/100 [00:07<00:00, 12.97it/s, loss=2278.1858] Epoch 12/20: 100%|██████████| 100/100 [00:07<00:00, 12.96it/s, loss=2090.5942] Epoch 13/20: 100%|██████████| 100/100 [00:07<00:00, 12.95it/s, loss=1939.2296] Epoch 14/20: 100%|██████████| 100/100 [00:07<00:00, 12.95it/s, loss=1813.9688] Epoch 15/20: 100%|██████████| 100/100 [00:07<00:00, 12.94it/s, loss=1708.0551] Epoch 16/20: 100%|██████████| 100/100 [00:07<00:00, 12.93it/s, loss=1628.9713] Epoch 17/20: 100%|██████████| 100/100 [00:07<00:00, 12.93it/s, loss=1545.6821] Epoch 18/20: 100%|██████████| 100/100 [00:07<00:00, 12.93it/s, loss=1917.6001] Epoch 19/20: 100%|██████████| 100/100 [00:07<00:00, 12.91it/s, loss=1416.0177] Epoch 20/20: 100%|██████████| 100/100 [00:07<00:00, 12.91it/s, loss=1359.5453] Process finished with exit code 0 ``` 训练完成后在输出目录（默认`./output`）下能看到生成的图片，默认每个epoch保存一次生成的图片。