InterLCM: Low-Quality Images as Intermediate States of Latent Consistency Models for Effective Blind Face Restoration

Official Gradio demo for Low-Quality Images as Intermediate States of Latent Consistency Models for Effective Blind Face Restoration (ICLR 2025)
🔥 InterLCM is a robust blind face restoration algorithm.
⭐ If InterLCM is helpful to your images or projects, please help star this repo. Thanks! 🤗

Input

Select InterLCM step (InterLCM enables 1-step⚡ BFR under non-extreme degradation conditions)

1 3

Pre_Face_Align

Background_Enhance

Face_Upsample

Output

Examples

If InterLCM is helpful, please help to ⭐ the Github Repo. Thanks!

📝 Citation If our work is useful for your research, please consider citing:

@inproceedings{li2025interlcm,
  title={InterLCM: Low-Quality Images as Intermediate States of Latent Consistency Models for Effective Blind Face Restoration},
  author={Li, Senmao and Wang, Kai and van de Weijer, Joost and Khan, Fahad Shahbaz and Guo, Chun-Le and Yang, Shiqi and Wang, Yaxing and Yang, Jian and Cheng, Ming-Ming},
  booktitle={ICLR},
  year={2025}
}

📧 Contact If you have any questions, please feel free to reach me out at senmaonk@gmail.com.