Research map / Image generation / Image restoration
Kernel Aware Resampler
Read the paperCheck 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 |
|---|---|---|---|
| Use kernel prediction with an estimated degradation map instead of SRWarp | KARL · ×2.0 | PSNR: 25.49 -> 27.39 | Helped |
| Use kernel prediction with an estimated degradation map instead of SRWarp | KARL · P×3 | PSNR: 22.70 -> 24.50 | Helped |
| Add sampling offsets to the resampling layer (ablation) | KARL · P×2 | PSNR: 24.77 -> 30.66 | Helped |
| Add sampling offsets to the resampling layer (ablation) | KARL · P×3 | PSNR: 22.68 -> 27.14 | Helped |
| Add the local Jacobian alongside sampling offsets (ablation) | KARL · P×2 | PSNR: 30.66 -> 30.95 | Helped |
| Add the local Jacobian alongside sampling offsets (ablation) | KARL · P×3 | PSNR: 27.14 -> 27.25 | Helped |
| Train with varied kernels rather than a fixed medium degradation (ablation) | KARL without degradation-map input · P×2 m | PSNR: 30.95 -> 29.69 | Hurt |
| Provide the degradation map when training with varied kernels (ablation) | KARL · P×2 s | PSNR: 29.61 -> 29.96 | Helped |
| Provide the degradation map when training with varied kernels (ablation) | KARL · P×2 m | PSNR: 29.69 -> 30.33 | Helped |
| Provide the degradation map when training with varied kernels (ablation) | KARL · P×2 l | PSNR: 24.98 -> 27.86 | Helped |
Closest papers by meaning
- Correction Filter for Single Image Super-Resolution: Robustifying Off-the-Shelf Deep Super-Resolvers (CVPR 2020 · similarity 0.89)
- Deep Constrained Least Squares for Blind Image Super-Resolution (CVPR 2022 · similarity 0.89)
- Blind Super-Resolution With Iterative Kernel Correction (CVPR 2019 · similarity 0.89)
- A Dynamic Kernel Prior Model for Unsupervised Blind Image Super-Resolution (CVPR 2024 · similarity 0.88)
- Unsupervised Degradation Representation Aware Transform for Real-World Blind Image Super-Resolution (AAAI 2025 · similarity 0.86)
- Unsupervised Diffusion-Based Degradation Modeling for Real-World Super-Resolution (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.