Papers
Topics
Authors
Recent
Gemini 2.5 Flash
Gemini 2.5 Flash
97 tokens/sec
GPT-4o
53 tokens/sec
Gemini 2.5 Pro Pro
44 tokens/sec
o3 Pro
5 tokens/sec
GPT-4.1 Pro
47 tokens/sec
DeepSeek R1 via Azure Pro
28 tokens/sec
2000 character limit reached

WaveFace: Authentic Face Restoration with Efficient Frequency Recovery (2403.12760v1)

Published 19 Mar 2024 in cs.CV

Abstract: Although diffusion models are rising as a powerful solution for blind face restoration, they are criticized for two problems: 1) slow training and inference speed, and 2) failure in preserving identity and recovering fine-grained facial details. In this work, we propose WaveFace to solve the problems in the frequency domain, where low- and high-frequency components decomposed by wavelet transformation are considered individually to maximize authenticity as well as efficiency. The diffusion model is applied to recover the low-frequency component only, which presents general information of the original image but 1/16 in size. To preserve the original identity, the generation is conditioned on the low-frequency component of low-quality images at each denoising step. Meanwhile, high-frequency components at multiple decomposition levels are handled by a unified network, which recovers complex facial details in a single step. Evaluations on four benchmark datasets show that: 1) WaveFace outperforms state-of-the-art methods in authenticity, especially in terms of identity preservation, and 2) authentic images are restored with the efficiency 10x faster than existing diffusion model-based BFR methods.

Definition Search Book Streamline Icon: https://streamlinehq.com
References (43)
  1. To learn image super-resolution, use a gan to learn how to do image degradation first. In ECCV, 2018.
  2. Progressive semantic-aware style transformation for blind face restoration. In CVPR, 2021.
  3. FSRNet: End-to-end learning face super-resolution with facial priors. In CVPR, 2018.
  4. Diffusion posterior sampling for general noisy inverse problems. In ICLR, 2023.
  5. Arcface: Additive angular margin loss for deep face recognition. In CVPR, 2019a.
  6. The menpo benchmark for multi-pose 2d and 3d facial landmark localisation and tracking. IJCV, 2019b.
  7. Diffusion models beat gans on image synthesis. In NeurIPS, 2021.
  8. Taming transformers for high-resolution image synthesis. In CVPR, 2021.
  9. VQFR: Blind face restoration with vector-quantized dictionary and parallel decoder. In ECCV, 2022.
  10. GANs trained by a two time-scale update rule converge to a local nash equilibrium. In NeurIPS, 2017.
  11. Denoising diffusion probabilistic models. In NeurIPS, 2020.
  12. Labeled faces in the wild: A database for studying face recognition in unconstrained environments. Tech. Report, 2008.
  13. Progressive growing of GANs for improved quality, stability, and variation. In ICLR, 2018.
  14. A style-based generator architecture for generative adversarial networks. In CVPR, 2019.
  15. Analyzing and improving the image quality of stylegan. In CVPR, 2020.
  16. Denoising diffusion restoration models. In NeurIPS, 2022.
  17. Adam: A method for stochastic optimization. In ICLR, 2015.
  18. Learning warped guidance for blind face restoration. In ECCV, 2018.
  19. SwinIR: Image restoration using swin transformer. In ICCV, 2021.
  20. Learning a no-reference quality metric for single-image super-resolution. CVIU, 2017.
  21. Making a “completely blind” image quality analyzer. SPL, 2012.
  22. Improved denoising diffusion probabilistic models. In ICML, 2021.
  23. Wavelet diffusion models are fast and scalable image generators. In CVPR, 2023.
  24. Diffusion autoencoders: Toward a meaningful and decodable representation. In CVPR, 2022.
  25. High-resolution image synthesis with latent diffusion models. In CVPR, 2022.
  26. Image super-resolution via iterative refinement. TPAMI, 2022.
  27. Deep semantic face deblurring. In CVPR, 2018.
  28. Denoising diffusion implicit models. arXiv:2010.02502, 2020.
  29. Attention is all you need. In NeurIPS, 2017.
  30. A survey of deep face restoration: Denoise, super-resolution, deblur, artifact removal. arXiv:2211.02831, 2022a.
  31. Towards real-world blind face restoration with generative facial prior. In CVPR, 2021a.
  32. Real-esrgan: Training real-world blind super-resolution with pure synthetic data. In ICCV Workshops, 2021b.
  33. Image quality assessment: from error visibility to structural similarity. TIP, 2004.
  34. Restoreformer: High-quality blind face restoration from undegraded key-value pairs. In CVPR, 2022b.
  35. DR2: Diffusion-based robust degradation remover for blind face restoration. In CVPR, 2023.
  36. Wider face: A face detection benchmark. In CVPR, 2016.
  37. Gan prior embedded network for blind face restoration in the wild. In CVPR, 2021.
  38. Face super-resolution guided by facial component heatmaps. In ECCV, 2018.
  39. Difface: Blind face restoration with diffused error contraction. arXiv:2212.06512, 2022.
  40. Edface-celeb-1 m: Benchmarking face hallucination with a million-scale dataset. TPAMI, 2022.
  41. The unreasonable effectiveness of deep features as a perceptual metric. In CVPR, 2018.
  42. Decoupled multi-task learning with cyclical self-regulation for face parsing. In CVPR, 2022.
  43. Towards robust blind face restoration with codebook lookup transformer. In NeurIPS, 2022.
User Edit Pencil Streamline Icon: https://streamlinehq.com
Authors (3)
  1. Yunqi Miao (5 papers)
  2. Jiankang Deng (96 papers)
  3. Jungong Han (111 papers)
Citations (2)

Summary

We haven't generated a summary for this paper yet.