Research map / Image generation / Image restoration
Degradation-Aware Metric Prompting for Hyperspectral Image Restoration
Binfeng Wang, Di Wang, Haonan Guo, Ying Fu, Jing Zhang
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 Degradation Prompts (DP) to the baseline (ablation) | DAMP baseline (no DP, no SSAM) · ARAD, average of 5 tasks | PSNR (dB): 45.82 -> 50.02 | Improved |
| add SSAM on top of DP (ablation) | DAMP · ARAD, average of 5 tasks | PSNR (dB): 50.02 -> 51.43 | Improved |
| route experts with DP instead of frequency-based routing (ablation) | DAMoE · ARAD, average of 5 tasks | PSNR (dB): 47.72 -> 51.43 | Improved |
| route with DP instead of degradation-type labels (ablation) | DAMoE · ARAD, average of 5 tasks | PSNR (dB): 46.27 -> 51.43 | Improved |
| route with DP instead of implicit black-box prompt (ablation) | DAMoE · ARAD, average of 5 tasks | PSNR (dB): 46.81 -> 51.43 | Improved |
| increase total experts from 2 to 4 (ablation) | DAMoE (top-1) · ARAD, average of 5 tasks | PSNR (dB): 49.97 -> 51.43 | Improved |
| increase total experts from 4 to 6 (ablation) | DAMoE (top-1) · ARAD, average of 5 tasks | PSNR (dB): 51.43 -> 51.24 | Hurt |
| activate more experts (1 to 4 of 4) (ablation) | DAMoE · ARAD, average of 5 tasks | PSNR (dB): 51.43 -> 50.01 | Hurt |
| add Gaussian noise (sigma 1.0) to routing logits (ablation) | DAMoE router · ARAD validation | PSNR (dB): 50.42 -> 51.43 | Improved |
| use 6 metrics instead of 3 in DP (ablation) | DP · ARAD, average of 5 tasks | PSNR (dB): 50.02 -> 51.43 | Improved |
| use 12 metrics instead of 6 in DP (ablation) | DP · ARAD, average of 5 tasks | PSNR (dB): 51.43 -> 51.51 | No clear effect |
| add HFER metric to DP (ablation) | DP · ARAD, average of 5 tasks | PSNR (dB): 50.18 -> 51.43 | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 2 of 2 on most shared results.
- MP-HSIR — MP-HSIR: A Multi-Prompt Framework for Universal Hyperspectral Image Restoration (ICCV 2025) · better on 4 of 4 shared results
- PromptIR — PromptIR: Prompting for All-in-One Image Restoration (NeurIPS 2023) · 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
- MP-HSIR: A Multi-Prompt Framework for Universal Hyperspectral Image Restoration (ICCV 2025 · similarity 0.91)
- Unsupervised Test-Time Adaptation Learning for Effective Hyperspectral Image Super-Resolution With Unknown Degeneration (TPAMI 2024 · similarity 0.89)
- A Trainable Spectral-Spatial Sparse Coding Model for Hyperspectral Image Restoration (NeurIPS 2021 · similarity 0.88)
- Unsupervised Adaptation Learning for Hyperspectral Imagery Super-Resolution (CVPR 2020 · similarity 0.88)
- Spectrum-Aware and Transferable Architecture Search for Hyperspectral Image Restoration (ECCV 2022 · similarity 0.87)
- HIR-Diff: Unsupervised Hyperspectral Image Restoration Via Improved Diffusion Models (CVPR 2024 · 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.
