Note: The original code is available at: https://github.com/junyanz/pytorch-CycleGAN-and-pix2pix.

Linux, Windows or macOS.

• Python 3, We use anaconda package.
• Matplotlib and OpenCV (to process images).
• Pytorch: https://pytorch.org/get-started/locally/.
• CPU or NVIDIA GPU + CUDA CuDNN.

• 1 Step: clone the original repository.

!git clone https://github.com/junyanz/pytorch-CycleGAN-and-pix2pix
cd pytorch-CycleGAN-and-pix2pix

• 2 Step: Install PyTorch and 0.4+ and other dependencies (e.g., torchvision, visdom and dominate).
  • For pip users, please type the command pip install -r requirements.txt.
  • For Conda users, you can create a new Conda environment using conda env create -f environment.yml.

• Pix2Pix:

  Create folder /path/to/data with subfolders A and B. A and B should each have their own subfolders train, val, test, etc. In /path/to/data/A/train, put training images in style A. In /path/to/data/B/train, put the corresponding images in style B. Repeat same for other data splits (val, test, etc).

  Corresponding images in a pair {A,B} must be the same size and have the same filename, e.g., /path/to/data/A/train/1.jpg is considered to correspond to /path/to/data/B/train/1.jpg.

  Once the data is formatted this way, call:

!python datasets/combine_A_and_B.py --fold_A /path/to/data/A --fold_B /path/to/data/B --fold_AB /path/to/data

This will combine each pair of images (A,B) into a single image file, ready for training.

• CycleGAN:

  Create a dataset folder under /dataset for your dataset.

  Create subfolders testA, testB, trainA, and trainB under your dataset's folder. Place any images you want to transform from a to b (cat2dog) in the testA folder, images you want to transform from b to a (dog2cat) in the testB folder, and do the same for the trainA and trainB folders.

!python datasets/make_dataset_aligned.py --dataset-path /media/pc/DATA/Gan-test/dataset_CycleGAN

• Pix2Pix:

  Change the --dataroot and --name to your own dataset's path and model's name. Use --gpu_ids 0,1,.. to train on multiple GPUs and --batch_size to change the batch size. Add --direction BtoA if you want to train a model to transfrom from class B to A.

!python train.py --dataroot /media/pc/DATA/Gan-test/ablacao/dataset_pix --name pix2pix-nyu --model pix2pix --direction AtoB --display_id -1

• CycleGAN:

!python train.py --dataroot /media/pc/DATA/Gan-test/data/database_CycleGAN_nyu_grey --name cyclegan-nyu --model cycle_gan --display_id -1 --continue_train --epoch_count 31

• Pix2Pix

!python test.py --dataroot /media/pc/DATA/Gan-test/data/database_CycleGAN_nyu_grey/testA --name pix2pix-nyu --model test --netG unet_256 --direction AtoB --dataset_mode single --norm batch

• CycleGAN

!python test.py --dataroot C:/Users/PC/GAN/pytorch-CycleGAN-nyu/teste --name cyclegan-nyu --model cycle_gan --phase test --no_dropout

• Training settings can be edited in: /options/train_options.py, base_options.py and test_options.py
• For changes to the model, simply modify the files: /models/cycle_gan_model.py and pix2pix_model.py

• (a) RGB input Image
• (b) Ground Truth Image
• (c) Depth Map Predicted by CycleGAN
• (d) Depth Map Predicted by Pix2Pix

To estimate depth maps using CycleGAN, it is important to use the reconstructed images in the prediction stage, and not the fake images generated by the network. The options can be observed in /models/cycle_gan_model.py.

@inproceedings{CycleGAN2017,
  title={Unpaired Image-to-Image Translation using Cycle-Consistent Adversarial Networks},
  author={Zhu, Jun-Yan and Park, Taesung and Isola, Phillip and Efros, Alexei A},
  booktitle={Computer Vision (ICCV), 2017 IEEE International Conference on},
  year={2017}
}


@inproceedings{isola2017image,
  title={Image-to-Image Translation with Conditional Adversarial Networks},
  author={Isola, Phillip and Zhu, Jun-Yan and Zhou, Tinghui and Efros, Alexei A},
  booktitle={Computer Vision and Pattern Recognition (CVPR), 2017 IEEE Conference on},
  year={2017}
}