WaveFace

Repository for "WaveFace: Authentic Face Restoration with Efficient Frequency Recovery" by Yunqi Miao, Jiankang Deng, and Jungong Han.

🔔 We are happy to announce that this work was accepted at CVPR 2024.

For more information, please check:

[Paper] [Project Page]

Evaluation

We release our synthetic testing dataset (CelebA_LQ_syn) and restoration results of CelebA-Test, LFW-Test, Wider-Test, WebPhoto-Test (ours).

Also, we provide data(CelebA_LQ_realesrgan_syn) & restoration results(Ours_RESR) of a synthetic dataset with RealESRGAN degradation.

If you find this project useful in your research, please consider citing:

@article{miao2024waveface,
  title={WaveFace: Authentic Face Restoration with Efficient Frequency Recovery},
  author={Miao, Yunqi and Deng, Jiankang and Han, Jungong},
  journal={arXiv preprint arXiv:2403.12760},
  year={2024}
}