Research map / Image generation / Super-resolution
Fill2SR: Repurposing Inpainting Diffusion Transformers for Real-World Super-Resolution
Xingfu Yi, Xiaoxue Yu
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 RCDT paired data (ablation) | Fill2SR · RealLQ250 (avg D1-D3) | MUSIQ: 59.9384 -> 67.7982 | Improved |
| add RCDT paired data (ablation) | Fill2SR · DIV2K_VAL (avg D1-D3) | LPIPS: 0.3866 -> 0.3981 | Hurt |
| add RCDT paired data (ablation) | Fill2SR · DIV2K_VAL (avg D1-D3) | PSNR: 22.4596 -> 21.7831 | Hurt |
| add RCDT paired data | Fill2SR · RealLQ250 4x | MUSIQ: 62.1914 -> 71.5498 | Improved |
| add Redux visual-semantic tokens (ablation) | Fill2SR · RealLQ250 (avg D1-D3) | MUSIQ: 67.1587 -> 67.7982 | Improved |
| use Redux tokens instead of noise-image tokens (ablation) | Fill2SR · DIV2K_VAL (avg D1-D3) | PSNR: 21.6229 -> 21.7831 | Improved |
| raise evidence strength alpha from 0.5 to 1 (ablation) | Fill2SR · DIV2K_VAL (avg D1-D3) | PSNR: 20.9871 -> 21.7831 | Improved |
| raise evidence strength alpha from 0.9 to 1 (ablation) | Fill2SR · DIV2K_VAL (avg D1-D3) | PSNR: 21.7308 -> 21.7831 | No clear effect |
| full generation (beta=1) instead of latent refinement beta=0.5 (ablation) | Fill2SR · RealLQ250 (avg D1-D3) | MUSIQ: 28.2379 -> 67.7982 | Improved |
| full generation (beta=1) instead of latent refinement beta=0.5 (ablation) | Fill2SR · DIV2K_VAL (avg D1-D3) | PSNR: 23.7536 -> 21.7831 | Hurt |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 3 of 3 on most shared results.
- DiT4SR — DiT4SR: Taming Diffusion Transformer for Real-World Image Super-Resolution (ICCV 2025) · better on 4 of 5 shared results
- FaithDiff — FaithDiff: Unleashing Diffusion Priors for Faithful Image Super-resolution (CVPR 2025) · better on 3 of 3 shared results
- SeeSR — SeeSR: Towards Semantics-Aware Real-World Image Super-Resolution (CVPR 2024) · better on 2 of 3 shared results
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
- When Latents Forget Pixels: Restoring Fidelity in Diffusion Transformer Super-Resolution (NeurIPS 2026 · similarity 0.89)
- DiT4SR: Taming Diffusion Transformer for Real-World Image Super-Resolution (ICCV 2025 · similarity 0.89)
- TSD-SR: One-Step Diffusion with Target Score Distillation for Real-World Image Super-Resolution (CVPR 2025 · similarity 0.88)
- RAD: Region-Aware Diffusion Models for Image Inpainting (CVPR 2025 · similarity 0.87)
- Edge-SD-SR: Low Latency and Parameter Efficient On-device Super-Resolution with Stable Diffusion via Bidirectional Conditioning (CVPR 2025 · similarity 0.87)
- One Diffusion Step to Real-World Super-Resolution via Flow Trajectory Distillation (ICML 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.
