Research map / Image generation / Image restoration
DynFaceRestore: Balancing Fidelity and Quality in Diffusion-Guided Blind Face Restoration with Dynamic Blur-Level Mapping and Guidance
Huu-Phu Do, Yu-Wei Chen, Yi-Cheng Liao, Chi-Wei Hsiao, Han-Yang Wang, Wei-Chen Chiu, Ching-Chun Huang
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 DBLM (blind input mapped to Gaussian blur, with diffusion guidance) (ablation) | diffusion baseline · CelebA-Test (ablation) | PSNR: 11.129 -> 24.992 | Improved |
| add DBLM (blind input mapped to Gaussian blur, with diffusion guidance) (ablation) | diffusion baseline · CelebA-Test (ablation) | FID: 55.777 -> 18.303 | Improved |
| add multiple Gaussian-blur guidance (ablation) | DBLM only (A) · CelebA-Test (ablation) | PSNR: 24.992 -> 25.094 | Improved |
| add multiple Gaussian-blur guidance (ablation) | DBLM only (A) · CelebA-Test (ablation) | FID: 18.303 -> 19.823 | Hurt |
| add multiple guidance (ablation) | DBLM+DSST+DGSA (E) · CelebA-Test (ablation) | PSNR: 24.327 -> 24.349 | Improved |
| add DSST starting-step table (ablation) | DBLM + multiple guidance (B) · CelebA-Test (ablation) | PSNR: 25.094 -> 25.107 | Improved |
| add DSST starting-step table (ablation) | DBLM + multiple guidance (B) · CelebA-Test (ablation) | FID: 19.823 -> 19.786 | Improved |
| add DSST starting-step table (ablation) | DBLM + multiple guidance (B) · CelebA-Test (ablation) | sampling step range: 1000 -> [690, 925] | Improved |
| add DGSA region-wise guidance scaling (ablation) | DBLM + multiple guidance (B) · CelebA-Test (ablation) | FID: 19.823 -> 15.007 | Improved |
| add DGSA region-wise guidance scaling (ablation) | DBLM + multiple guidance + DSST (C) · CelebA-Test (ablation) | PSNR: 25.107 -> 24.349 | Hurt |
| add DGSA region-wise guidance scaling (ablation) | DBLM + multiple guidance + DSST (C) · CelebA-Test (ablation) | FID: 19.786 -> 14.780 | Improved |
| use kernel-convolved RM(y) instead of TM output as DBLM output (ablation) | DynFaceRestore · CelebA-Test | PSNR: 24.261 -> 24.349 | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 2 on most shared results.
- DifFace — DifFace: Blind Face Restoration With Diffused Error Contraction (TPAMI 2024) · better on 2 of 3 shared results
- RestoreFormer — RestoreFormer: High-Quality Blind Face Restoration From Undegraded Key-Value Pairs (CVPR 2022) · better on 0 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
- Self-Supervised Selective-Guided Diffusion Model for Old-Photo Face Restoration (NeurIPS 2025 · similarity 0.88)
- Blind Face Restoration under Extreme Conditions: Leveraging 3D-2D Prior Fusion for Superior Structural and Texture Recovery (AAAI 2024 · similarity 0.88)
- $InterLCM$: Low-Quality Images as Intermediate States of Latent Consistency Models for Effective Blind Face Restoration (ICLR 2025 · similarity 0.88)
- FaceMe: Robust Blind Face Restoration with Personal Identification (AAAI 2025 · similarity 0.88)
- OSDFace: One-Step Diffusion Model for Face Restoration (CVPR 2025 · similarity 0.87)
- Overcoming False Illusions in Real-World Face Restoration with Multi-Modal Guided Diffusion Model (ICLR 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.
