EleGANt: Exquisite and Locally Editable GAN for Makeup Transfer

Updated dependencies and refactored legacy code to be compatible with the latest environment.

Custom PyTorch implementation of ECCV 2022 paper "EleGANt: Exquisite and Locally Editable GAN for Makeup Transfer"

Chenyu Yang, Wanrong He, Yingqing Xu, and Yang Gao.

Getting Started

Installation

This code was tested on Windows 11 with CUDA 13.2.

a. Create a virtual environment and activate it.

python -m venv venv
venv\Scripts\activate
python -m pip install --upgrade pip

b. Install PyTorch and torchvision.

pip install torch torchvision

c. Install other required libaries.

pip install --upgrade setuptools wheel
pip install opencv-python matplotlib dlib fvcore

• Prepare Dataset & Checkpoints

Test

To test our model, download the weights of the trained model and run

python scripts/demo.py

Examples of makeup transfer results can be seen here.

Train

To train a model from scratch, run

python scripts/train.py

Customized Transfer

This is our demo of customized makeup editing. The interactive system is built upon Streamlit and the interface in ./training/inference.py.

Controllable makeup transfer.

Local makeup editing.

Citation

If this work is helpful for your research, please consider citing the following BibTeX entry.

@article{yang2022elegant,
  title={EleGANt: Exquisite and Locally Editable GAN for Makeup Transfer},
  author={Yang, Chenyu and He, Wanrong and Xu, Yingqing and Gao, Yang}
  journal={arXiv preprint arXiv:2207.09840},
  year={2022}
}

Acknowledgement

Some of the codes are build upon PSGAN and aster.Pytorch.

License

This work is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License.