Research map / Image generation / Image restoration
Towards Real-World Blind Face Restoration With Generative Facial Prior
Check your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 multi-resolution spatial modulation to latent code mapping (ablation) | GFP-GAN · CelebA-Test blind face restoration | LPIPS: 0.550 ($\uparrow$) -> 0.3646 | Helped |
| Use CS-SFT instead of standard SFT (ablation) | GFP-GAN · CelebA-Test blind face restoration | LPIPS: 0.387 ($\uparrow$) -> 0.3646 | Helped |
| Add pretrained generative facial prior (ablation) | GFP-GAN · CelebA-Test blind face restoration | LPIPS: 0.379 ($\uparrow$) -> 0.3646 | Helped |
| Add pyramid restoration loss (ablation) | GFP-GAN · CelebA-Test blind face restoration | LPIPS: 0.369 ($\uparrow$) -> 0.3646 | Helped |
| Use CS-SFT instead of standard SFT (ablation) | GFP-GAN · CelebA-Test blind face restoration | FID: 47.65 ($\uparrow$) -> 42.62 | Helped |
| Replace full CS-SFT modulation with standard SFT (ablation) | GFP-GAN · CelebA-Test blind face restoration | identity distance (Deg.): 34.60 -> 34.38 ($\downarrow$) | Mixed |
| Add facial component discriminators and feature style loss (ablation) | GFP-GAN · CelebA-Test blind face restoration | facial detail quality | Helped |
| Use Gram-statistics feature style loss rather than feature matching loss (ablation) | GFP-GAN facial component loss · CelebA-Test blind face restoration | facial detail quality | Helped |
Closest papers by meaning
- GAN Prior Embedded Network for Blind Face Restoration in the Wild (CVPR 2021 · similarity 0.92)
- Blind Face Restoration under Extreme Conditions: Leveraging 3D-2D Prior Fusion for Superior Structural and Texture Recovery (AAAI 2024 · similarity 0.89)
- In-N-Out: Faithful 3D GAN Inversion with Volumetric Decomposition for Face Editing (CVPR 2024 · similarity 0.87)
- OSDFace: One-Step Diffusion Model for Face Restoration (CVPR 2025 · similarity 0.87)
- Learning Degradation-unaware Representation with Prior-based Latent Transformations for Blind Face Restoration (CVPR 2024 · similarity 0.86)
- FaceMe: Robust Blind Face Restoration with Personal Identification (AAAI 2025 · similarity 0.86)
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.