CycleGAN and pix2pix in PyTorch

This is our PyTorch implementation for both unpaired and paired image-to-image translation. It is still under active development.

The code was written by Jun-Yan Zhu and Taesung Park, and supported by Tongzhou Wang.

This PyTorch implementation produces results comparable to or better than our original Torch software. If you would like to reproduce the exact same results as in the papers, check out the original CycleGAN Torch and pix2pix Torch code

Note: The current software works well with PyTorch 0.4. Check out the older branch that supports PyTorch 0.1-0.3.

CycleGAN: [Project] [Paper] [Torch]

Pix2pix: [Project] [Paper] [Torch]

[EdgesCats Demo] [pix2pix-tensorflow]

Written by Christopher Hesse

If you use this code for your research, please cite:

Unpaired Image-to-Image Translation using Cycle-Consistent Adversarial Networks Jun-Yan Zhu*, Taesung Park*, Phillip Isola, Alexei A. Efros In ICCV 2017. (* equal contributions)

Image-to-Image Translation with Conditional Adversarial Networks Phillip Isola, Jun-Yan Zhu, Tinghui Zhou, Alexei A. Efros In CVPR 2017.

Course

CycleGAN course assignment code and handout designed by Prof. Roger Grosse for CSC321 "Intro to Neural Networks and Machine Learning" at University of Toronto. Please contact the instructor if you would like to adopt it in your course.

Other implementations

CycleGAN

[Tensorflow] (by Harry Yang), [Tensorflow] (by Archit Rathore), [Tensorflow] (by Van Huy), [Tensorflow] (by Xiaowei Hu), [Tensorflow-simple] (by Zhenliang He), [TensorLayer] (by luoxier), [Chainer] (by Yanghua Jin), [Minimal PyTorch] (by yunjey), [Mxnet] (by Ldpe2G), [lasagne/keras] (by tjwei)

pix2pix

[Tensorflow] (by Christopher Hesse), [Tensorflow] (by Eyyüb Sariu), [Tensorflow (face2face)] (by Dat Tran), [Tensorflow (film)] (by Arthur Juliani), [Tensorflow (zi2zi)] (by Yuchen Tian), [Chainer] (by mattya), [tf/torch/keras/lasagne] (by tjwei), [Pytorch] (by taey16)

Prerequisites

• Linux or macOS
• Python 2 or 3
• CPU or NVIDIA GPU + CUDA CuDNN

Getting Started

Installation

• Install PyTorch 0.4 and dependencies from http://pytorch.org
• Install Torch vision from the source.

git clone https://github.com/pytorch/vision
cd vision
python setup.py install

• Install python libraries visdom and dominate.

pip install visdom
pip install dominate

• Alternatively, all dependencies can be installed by

pip install -r requirements.txt

• Clone this repo:

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

CycleGAN train/test

• Download a CycleGAN dataset (e.g. maps):

bash ./datasets/download_cyclegan_dataset.sh maps

• Train a model:

#!./scripts/train_cyclegan.sh
python train.py --dataroot ./datasets/maps --name maps_cyclegan --model cycle_gan

• To view training results and loss plots, run python -m visdom.server and click the URL http://localhost:8097. To see more intermediate results, check out ./checkpoints/maps_cyclegan/web/index.html
• Test the model:

#!./scripts/test_cyclegan.sh
python test.py --dataroot ./datasets/maps --name maps_cyclegan --model cycle_gan

The test results will be saved to a html file here: ./results/maps_cyclegan/latest_test/index.html.

pix2pix train/test

• Download a pix2pix dataset (e.g.facades):

bash ./datasets/download_pix2pix_dataset.sh facades

• Train a model:

#!./scripts/train_pix2pix.sh
python train.py --dataroot ./datasets/facades --name facades_pix2pix --model pix2pix --which_direction BtoA

• To view training results and loss plots, run python -m visdom.server and click the URL http://localhost:8097. To see more intermediate results, check out ./checkpoints/facades_pix2pix/web/index.html
• Test the model (bash ./scripts/test_pix2pix.sh):

#!./scripts/test_pix2pix.sh
python test.py --dataroot ./datasets/facades --name facades_pix2pix --model pix2pix --which_direction BtoA 

The test results will be saved to a html file here: ./results/facades_pix2pix/test_latest/index.html.

More example scripts can be found at scripts directory.

Apply a pre-trained model (CycleGAN)

• You can download a pretrained model (e.g. horse2zebra) with the following script:

bash pretrained_models/download_cyclegan_model.sh horse2zebra

The pretrained model is saved at ./checkpoints/{name}_pretrained/latest_net_G.pth.

• To test the model, you also need to download the horse2zebra dataset:

bash ./datasets/download_cyclegan_dataset.sh horse2zebra

• Then generate the results using

python test.py --dataroot datasets/horse2zebra/testA --name horse2zebra_pretrained --model test

The option --model test is used for generating results of CycleGAN only for one side. python test.py --model cycle_gan will require loading and generating results in both directions, which is sometimes unnecessary. The results will be saved at ./results/. Use --results_dir {directory_path_to_save_result} to specify the results directory.

• Note: The models trained using Torch and PyTorch produce slightly different results, although we were not able to decide which result is better. If you would like to reproduce the same results in our paper, we recommend using the pretrained models in the Torch codebase.

• If you would like to apply a pre-trained model to a collection of input images (rather than image pairs), please use --dataset_mode single and --model test options. Here is a script to apply a model to Facade label maps (stored in the directory facades/testB).

#!./scripts/test_single.sh
python test.py --dataroot ./datasets/facades/testB/ --name {your_trained_model_name} --model test 

You might want to specify --which_model_netG to match the generator architecture of the trained model.

Apply a pre-trained model (pix2pix)

Download a pre-trained model with ./pretrained_models/download_pix2pix_model.sh.

• For example, if you would like to download label2photo model on the Facades dataset,

bash pretrained_models/download_pix2pix_model.sh facades_label2photo

• Download the pix2pix facades datasets

bash ./datasets/download_pix2pix_dataset.sh facades

• Then generate the results using

python test.py --dataroot ./datasets/facades/ --which_direction BtoA --model pix2pix --name facades_label2photo_pretrained

Note that we specified --which_direction BtoA as Facades dataset's A to B direction is photos to labels.

• See a list of currently available models at bash pretrained_models/download_pix2pix_model.sh

Training/test Details

• Flags: see options/train_options.py and options/base_options.py for the training flags; see options/test_options.py and options/base_options.py for the test flags. There are some model-specific flags as well, which are added in the model files, such as --lambda_A option in model/cycle_gan_model.py. The default values of these options are also adjusted in the model files.
• CPU/GPU (default --gpu_ids 0): set--gpu_ids -1 to use CPU mode; set --gpu_ids 0,1,2 for multi-GPU mode. You need a large batch size (e.g. --batchSize 32) to benefit from multiple GPUs.
• Visualization: during training, the current results can be viewed using two methods. First, if you set --display_id > 0, the results and loss plot will appear on a local graphics web server launched by visdom. To do this, you should have visdom installed and a server running by the command python -m visdom.server. The default server URL is http://localhost:8097. display_id corresponds to the window ID that is displayed on the visdom server. The visdom display functionality is turned on by default. To avoid the extra overhead of communicating with visdom set --display_id -1. Second, the intermediate results are saved to [opt.checkpoints_dir]/[opt.name]/web/ as an HTML file. To avoid this, set --no_html.
• Preprocessing: images can be resized and cropped in different ways using --resize_or_crop option. The default option 'resize_and_crop' resizes the image to be of size (opt.loadSize, opt.loadSize) and does a random crop of size (opt.fineSize, opt.fineSize). 'crop' skips the resizing step and only performs random cropping. 'scale_width' resizes the image to have width opt.fineSize while keeping the aspect ratio. 'scale_width_and_crop' first resizes the image to have width opt.loadSize and then does random cropping of size (opt.fineSize, opt.fineSize).
• Fine-tuning/Resume training: to fine-tune a pre-trained model, or resume the previous training, use the --continue_train flag. The program will then load the model based on which_epoch. By default, the program will initialize the epoch count as 1. Set --epoch_count <int> to specify a different starting epoch count.
• For Conda users, we include a script ./scripts/conda_deps.sh to install PyTorch and other libraries.

CycleGAN Datasets

Download the CycleGAN datasets using the following script. Some of the datasets are collected by other researchers. Please cite their papers if you use the data.

bash ./datasets/download_cyclegan_dataset.sh dataset_name

• facades: 400 images from the CMP Facades dataset. [Citation]
• cityscapes: 2975 images from the Cityscapes training set. [Citation]
• maps: 1096 training images scraped from Google Maps.
• horse2zebra: 939 horse images and 1177 zebra images downloaded from ImageNet using keywords wild horse and zebra
• apple2orange: 996 apple images and 1020 orange images downloaded from ImageNet using keywords apple and navel orange.
• summer2winter_yosemite: 1273 summer Yosemite images and 854 winter Yosemite images were downloaded using Flickr API. See more details in our paper.
• monet2photo, vangogh2photo, ukiyoe2photo, cezanne2photo: The art images were downloaded from Wikiart. The real photos are downloaded from Flickr using the combination of the tags landscape and landscapephotography. The training set size of each class is Monet:1074, Cezanne:584, Van Gogh:401, Ukiyo-e:1433, Photographs:6853.
• iphone2dslr_flower: both classes of images were downlaoded from Flickr. The training set size of each class is iPhone:1813, DSLR:3316. See more details in our paper.

To train a model on your own datasets, you need to create a data folder with two subdirectories trainA and trainB that contain images from domain A and B. You can test your model on your training set by setting --phase train in test.py. You can also create subdirectories testA and testB if you have test data.

You should not expect our method to work on just any random combination of input and output datasets (e.g. cats<->keyboards). From our experiments, we find it works better if two datasets share similar visual content. For example, landscape painting<->landscape photographs works much better than portrait painting <-> landscape photographs. zebras<->horses achieves compelling results while cats<->dogs completely fails.

pix2pix datasets

Download the pix2pix datasets using the following script. Some of the datasets are collected by other researchers. Please cite their papers if you use the data.

bash ./datasets/download_pix2pix_dataset.sh dataset_name

• facades: 400 images from CMP Facades dataset. [Citation]
• cityscapes: 2975 images from the Cityscapes training set. [Citation]
• maps: 1096 training images scraped from Google Maps
• edges2shoes: 50k training images from UT Zappos50K dataset. Edges are computed by HED edge detector + post-processing. [Citation]
• edges2handbags: 137K Amazon Handbag images from iGAN project. Edges are computed by HED edge detector + post-processing. [Citation]

We provide a python script to generate pix2pix training data in the form of pairs of images {A,B}, where A and B are two different depictions of the same underlying scene. For example, these might be pairs {label map, photo} or {bw image, color image}. Then we can learn to translate A to B or B to A:

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.

Citation

If you use this code for your research, please cite our papers.

@inproceedings{CycleGAN2017,
  title={Unpaired Image-to-Image Translation using Cycle-Consistent Adversarial Networkss},
  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}
}

Related Projects

CycleGAN: Unpaired Image-to-Image Translation using Cycle-Consistent Adversarial Networks pix2pix: Image-to-image translation with conditional adversarial nets iGAN: Interactive Image Generation via Generative Adversarial Networks

Cat Paper Collection

If you love cats, and love reading cool graphics, vision, and learning papers, please check out the Cat Paper Collection: [Github] [Webpage]

Acknowledgments

Code is inspired by pytorch-DCGAN.