Research map / Image generation / Image restoration
PRISM: Controllable Diffusion for Compound Image Restoration with Scientific Fidelity
Rupa Kurinchi-Vendhan, Pratyusha Sharma, Antonio Torralba, Sara Beery
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 |
|---|---|---|---|
| use selective instead of full restoration | PRISM · Remote sensing downstream | accuracy: 0.842 -> 0.836 | No clear effect |
| use selective instead of full restoration | PRISM · Camera traps downstream | accuracy: 0.976 -> 0.984 | Improved |
| use selective instead of full restoration | PRISM · Microscopy downstream | mIoU: 0.475 -> 0.580 | Improved |
| use selective instead of full restoration | PRISM · Urban scenes downstream | mIoU: 0.615 -> 0.650 | Improved |
| add denoising on top of super-resolution | PRISM · BioSR segmentation | mIoU: 0.580 -> 0.475 | Hurt |
| restore with super-resolution instead of denoising | PRISM · BioSR fluorescence | MSE: 0.006 -> 0.018 | Hurt |
| restore with super-resolution instead of no restoration | PRISM · BioSR segmentation | mIoU: 0.356 -> 0.580 | Improved |
| train on compound degradations instead of primitives only (ablation) | PRISM · MDB (1 to 4 distortions) | PSNR | Improved |
| add compound-aware contrastive CLIP fine-tuning (ablation) | PRISM · MDB sequential vs composite prompting | PSNR | Improved |
| use PRISM instead of MPerceiver | MDB · MDB | FID: 48.18 -> 48.97 | Hurt |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 4 of 4 on most shared results.
- AutoDIR — AutoDIR: Automatic All-in-One Image Restoration with Latent Diffusion (ECCV 2024) · better on 5 of 5 shared results
- DiffPlugin — Diff-Plugin: Revitalizing Details for Diffusion-based Low-level Tasks (CVPR 2024) · better on 2 of 2 shared results
- OneRestore — OneRestore: A Universal Restoration Framework for Composite Degradation (ECCV 2024) · better on 2 of 2 shared results
- PromptIR — PromptIR: Prompting for All-in-One Image Restoration (NeurIPS 2023) · better on 2 of 2 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
- Discovering Unseen Degradations to Adapt Open-World Image Restoration (NeurIPS 2026 · similarity 0.90)
- UniRes: Universal Image Restoration for Complex Degradations (ICCV 2025 · similarity 0.89)
- OneRestore: A Universal Restoration Framework for Composite Degradation (ECCV 2024 · similarity 0.89)
- A Survey on All-in-One Image Restoration: Taxonomy, Evaluation and Future Trends (TPAMI 2025 · similarity 0.88)
- Learning Where and What to Restore for Composite Image Restoration (NeurIPS 2026 · similarity 0.88)
- Visual-Instructed Degradation Diffusion for All-in-One Image Restoration (CVPR 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.
