AsymmetricGAN_multi

Dependencies

• Python 3.5+
• PyTorch 0.4.0+
• TensorFlow 1.3+ (optional for tensorboard)

Usage

1. Cloning the repository

$ git https://github.com/Ha0Tang/AsymmetricGAN
$ cd AsymmetricGAN_multi/

2. Downloading the dataset

To download the CelebA dataset:

$ bash download.sh celeba

To download the RaFD dataset, you must request access to the dataset from the Radboud Faces Database website. Then, you need to create a folder structure as described here.

3. Training

$ bash train_asymmetricgan.sh

4. Testing

$ bash test_asymmetricgan.sh

5. Pretrained modelTo download a pretrained model checkpoint, run the script below.

$ bash download_pretrained.sh rafd_generator1
$ bash download_pretrained.sh rafd_generator2
$ bash download_pretrained.sh rafd_generator3

To translate images using the pretrained model, run the evaluation script below.

$ python main.py --mode test --dataset RaFD --image_size 256 --c_dim 8 \
                 --rafd_image_dir data/RaFD/test \
                 --model_save_dir rafd_generator1_pretrained/models \
                 --result_dir rafd_generator1_pretrained/results