Research map / Image generation / Image restoration
Interactive Multi-Dimension Modulation with Dynamic Controllable Residual Learning for Image Restoration
Jingwen He, Chao Dong, Yu Qiao
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 global controllable residual connection (ablation) | CResMD · CBSD68 blur r0.5 noise sigma0 | PSNR: 52.70 -> 53.17 | Improved |
| add global controllable residual connection (ablation) | CResMD · CBSD68 blur r1 noise sigma0 | PSNR: 38.04 -> 38.38 | Improved |
| add global controllable residual connection (ablation) | CResMD · LIVE1 blur r0.5 noise sigma0 | PSNR: 51.22 -> 52.21 | Improved |
| add global controllable residual connection (ablation) | CResMD · CBSD68 blur r0 noise sigma0 (identity) | PSNR: 71.39 -> +inf | Improved |
| increase number of local controllable residual connections from 1 to 32 (ablation) | CResMD · CBSD68 2D modulation | PSNR distance: sharp leap 0.22dB in deblurring r1 from 4 to 8 connections | Improved |
| increase local connections from 1 to 32 on denoising (ablation) | CResMD · CBSD68 denoising sigma15 and sigma50 | PSNR distance: less than 0.1dB | No clear effect |
| use beta sampling alpha=0.5 beta=1.0 instead of uniform (ablation) | CResMD · LIVE1 sampling curve | total PSNR: +0.28 | Improved |
| use beta sampling alpha=0.2 beta=1.0 instead of uniform (ablation) | CResMD · LIVE1 sampling curve | total PSNR: -0.50 | Hurt |
| use beta sampling alpha=1.0 beta=2.0 instead of uniform (ablation) | CResMD · LIVE1 sampling curve | total PSNR: +0.07 | Mixed |
| use CResMD instead of DNI/AdaFM/CFSNet on single-degradation deblurring | CResMD · CBSD68 deblurring r2 to r4 | PSNR distance: >1.0 dB | Improved |
Closest papers by meaning
- Interactive Multi-Dimension Modulation for Image Restoration (TPAMI 2021 · similarity 0.98)
- Empowering Real-World Image Super-Resolution With Flexible Interactive Modulation (TPAMI 2024 · similarity 0.89)
- UP-Restorer: When Unrolling Meets Prompts for Unified Image Restoration (AAAI 2025 · similarity 0.85)
- Residual-Conditioned Optimal Transport: Towards Structure-Preserving Unpaired and Paired Image Restoration (ICML 2024 · similarity 0.85)
- Learning Where and What to Restore for Composite Image Restoration (NeurIPS 2026 · similarity 0.85)
- AdaIR: Adaptive All-in-One Image Restoration via Frequency Mining and Modulation (ICLR 2025 · similarity 0.85)
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.
