StarGAN-PyTorch

Contents

• Introduction
• Getting Started
  • Requirements
  • Install
• All pretrained model weights
• Test (e.g. CelebA-128x128)
• Train
• Contributing
• Credit
  • StarGAN: Unified Generative Adversarial Networks for Multi-Domain Image-to-Image Translation

Introduction

This repository contains an op-for-op PyTorch reimplementation of StarGAN: Unified Generative Adversarial Networks for Multi-Domain Image-to-Image Translation.

Getting Started

Requirements

• Python 3.10+
• PyTorch 2.1.0+
• CUDA 11.8+
• Ubuntu 22.04+

Install

git clone https://github.com/Lornatang/StarGAN-PyTorch.git
cd StarGAN-PyTorch
pip install torch torchvision torchaudio --index-url https://download.pytorch.org/whl/cu118
pip install -r requirements.txt
# From pypi (recommended)
pip install stargan_pytorch
# From local
pip install -e .

All pretrained model weights

• g_128x128-celeba
• g_256x256-celeba
• d_128x128-celeba
• d_256x256-celeba

Test (e.g. CelebA-128x128)

# Download g_128x128-celeba model weights to `./results/pretrained_models`
wget https://huggingface.co/goodfellowliu/StarGAN-PyTorch/resolve/main/g_128x128-celeba.pth.tar?download=true -O ./results/pretrained_models/g_128x128-celeba.pth.tar
python ./tools/eval.py ./configs/celeba_128x128.yaml
# Result will be saved to `./results/test/celeba_128x128`

original - Black_Hair - Blond_Hair - Brown_Hair - Male - Young

Train

Please refer to README.md in the data directory for the method of making a dataset.

# If you want to train g_128x128-celeba, run this command
python ./tools/train.py ./configs/celeba_128x128.yaml
# If you want to train g_256x256-celeba, run this command
python ./tools/train.py ./configs/celeba_256x256.yaml

The training results will be saved to ./results/train/celeba_128x128 or ./results/train/celeba_256x256.

Contributing

If you find a bug, create a GitHub issue, or even better, submit a pull request. Similarly, if you have questions, simply post them as GitHub issues.

I look forward to seeing what the community does with these models!

Credit

StarGAN: Unified Generative Adversarial Networks for Multi-Domain Image-to-Image Translation

Yunjey Choi, Minje Choi, Munyoung Kim, Jung-Woo Ha, Sunghun Kim, Jaegul Choo

Abstract
Recent studies have shown remarkable success in imageto-image translation for two domains. However, existing approaches have limited scalability and robustness in handling more than two domains, since different models should be built independently for every pair of image domains. To address this limitation, we propose StarGAN, a novel and scalable approach that can perform image-to-image translations for multiple domains using only a single model. Such a unified model architecture of StarGAN allows simultaneous training of multiple datasets with different domains within a single network. This leads to StarGAN’s superior quality of translated images compared to existing models as well as the novel capability of flexibly translating an input image to any desired target domain. We empirically demonstrate the effectiveness of our approach on a facial attribute transfer and a facial expression synthesis tasks.

[Paper] [Code(PyTorch)]

@misc{choi2018stargan,
      title={StarGAN: Unified Generative Adversarial Networks for Multi-Domain Image-to-Image Translation}, 
      author={Yunjey Choi and Minje Choi and Munyoung Kim and Jung-Woo Ha and Sunghun Kim and Jaegul Choo},
      year={2018},
      eprint={1711.09020},
      archivePrefix={arXiv},
      primaryClass={cs.CV}
}