Research map / Image generation / Editing and style transfer
Learning Warped Guidance for Blind Face Restoration
Xiaoming Li, Ming Liu, Yuting Ye, Wangmeng Zuo, Liang Lin, Ruigang Yang
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| add WarpNet (warped guidance) instead of concatenating guide image (ablation) | GFRNet · WebFace test (4x) | PSNR (dB): 29.66 -> 32.31 | Improved |
| add flow loss (landmark loss and TV regularization) to WarpNet training (ablation) | GFRNet · WebFace test (4x) | PSNR (dB): 31.43 -> 32.31 | Improved |
| add flow loss to WarpNet training (ablation) | GFRNet · VggFace2 test (4x) | PSNR (dB): 26.61 -> 28.55 | Improved |
| add guided image input (ablation) | GFRNet without WarpNet · VggFace2 test (4x) | PSNR (dB): 25.97 -> 26.03 | Improved |
| use guided image of same identity instead of random identity (ablation) | GFRNet · WebFace test (4x) | PSNR (dB): 31.46 -> 32.31 | Improved |
| add flow loss when WarpNet is present vs no WarpNet (ablation) | GFRNet · VggFace2 test (8x) | PSNR (dB): 23.29 vs 23.17 | Mixed |
| double channels of no-WarpNet variant to match parameter count (ablation) | Ours(-W) · VggFace2 test (4x) | PSNR (dB): 26.03 -> 27.25 | Improved |
| add WarpNet and flow loss over parameter-matched variant (ablation) | Ours(-W2) · WebFace test (4x) | PSNR (dB): 29.73 -> 32.31 | Improved |
| retrain DeblurGAN on own data with guided image input (MDeblurGAN) | DeblurGAN · VggFace2 test (4x) | PSNR (dB): 24.65 -> 25.32 | Improved |
| retrain DnCNN on own data with guided image input (MDnCNN) | DnCNN · VggFace2 test (4x) | PSNR (dB): 26.73 -> 27.05 | Improved |
| retrain ARCNN on own data with guided image input (MARCNN) | ARCNN · VggFace2 test (4x) | PSNR (dB): 22.05 -> 25.43 | Improved |
| use GFRNet instead of retrained baselines | MDnCNN · WebFace test (4x) | PSNR (dB): 29.40 -> 32.31 | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 2 of 2 on most shared results.
- WaveletSRNet — Wavelet-SRNet: A Wavelet-Based CNN for Multi-Scale Face Super Resolution (ICCV 2017) · better on 1 of 1 shared result
- MemNet — MemNet: A Persistent Memory Network for Image Restoration (ICCV 2017) · better on 1 of 1 shared result
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- RestoreFormer++: Towards Real-World Blind Face Restoration From Undegraded Key-Value Pairs (TPAMI 2023 · similarity 0.89)
- A Natural Language Guided Approach for Blind Face Restoration: Methodology and Dataset (TPAMI 2026 · similarity 0.89)
- Learning Degradation-unaware Representation with Prior-based Latent Transformations for Blind Face Restoration (CVPR 2024 · similarity 0.88)
- Blind Face Restoration under Extreme Conditions: Leveraging 3D-2D Prior Fusion for Superior Structural and Texture Recovery (AAAI 2024 · similarity 0.87)
- $InterLCM$: Low-Quality Images as Intermediate States of Latent Consistency Models for Effective Blind Face Restoration (ICLR 2025 · similarity 0.87)
- FaceMe: Robust Blind Face Restoration with Personal Identification (AAAI 2025 · similarity 0.87)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
