Research map / Image generation / Super-resolution
Learning Heterogeneous Degradation Representation for Real-World Super-Resolution
Haowei Li, Pengxu Wei, Dongyu Zhang, Liang Lin
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 channel modulation (posterior mode guidance) (ablation) | Ours SR network · DRealSR (x4) | PSNR: 27.78 -> 28.27 | Improved |
| add channel modulation (ablation) | Ours SR network · RealSR (x4) | LPIPS: 0.2832 -> 0.2455 | Improved |
| add channel modulation (ablation) | Ours SR network · SVSR (x4) | PSNR: 24.70 -> 25.13 | Improved |
| add spatial modulation (posterior variance guidance) (ablation) | Ours SR network · SVSR (x4) | LPIPS: 0.3360 -> 0.3181 | Improved |
| add spatial modulation (ablation) | Ours SR network · DRealSR (x4) | PSNR: 28.24 -> 28.27 | No clear effect |
| add spatial modulation (ablation) | Ours SR network · DIV2K (x4) | PSNR: 20.01 -> 20.03 | No clear effect |
| add SAVL (probabilistic MI-suppressed representation) over deterministic code (ablation) | degradation representation · SVSR linear probe, Scene-ID accuracy | accuracy (lower better): 0.9404 -> 0.4997 | Improved |
| add SAVL over deterministic code (ablation) | degradation representation · SVSR linear probe, noise-level accuracy | accuracy: 0.8148 -> 0.9480 | Improved |
| add SAVL over deterministic code plus CLUB bound (ablation) | degradation representation · SVSR linear probe, noise-level accuracy | accuracy: 0.2000 -> 0.9480 | Improved |
| add SAVL over MANet (ablation) | degradation representation · Spatially variant degradation type clustering | accuracy (%): 43.66% -> 72.44% | Improved |
| add SAVL over deterministic code (ablation) | degradation representation · Spatially variant blur level clustering | accuracy (%): 39.34% -> 61.20% | Improved |
| add SAVL over deterministic code (ablation) | degradation representation · Spatially variant noise strength clustering | accuracy (%): 51.36% -> 57.60% | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- LightBSR — LightBSR: Towards Lightweight Blind Super-Resolution via Discriminative Implicit Degradation Representation Learning (ICCV 2025) · better on 3 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
- Learning Controllable Degradation for Real-World Super-Resolution via Constrained Flows (ICML 2023 · similarity 0.91)
- Unsupervised Diffusion-Based Degradation Modeling for Real-World Super-Resolution (AAAI 2025 · similarity 0.90)
- Unsupervised Degradation Representation Aware Transform for Real-World Blind Image Super-Resolution (AAAI 2025 · similarity 0.89)
- DORNet: A Degradation Oriented and Regularized Network for Blind Depth Super-Resolution (CVPR 2025 · similarity 0.88)
- Low-Res Leads the Way: Improving Generalization for Super-Resolution by Self-Supervised Learning (CVPR 2024 · similarity 0.88)
- Towards Realistic Data Generation for Real-World Super-Resolution (ICLR 2025 · similarity 0.88)
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.
