Research map / Image generation / Image restoration
A Survey on All-in-One Image Restoration: Taxonomy, Evaluation and Future Trends
Junjun Jiang, Zengyuan Zuo, Gang Wu, Kui Jiang, Xianming Liu
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 all-in-one training instead of a general multi-task model (PromptIR vs Restormer) | Restormer-class backbone · Three-degradation average | PSNR: 30.75 -> 32.06 | Improved |
| add prompt-in-prompt learning to Restormer | Restormer · Three-degradation average | PSNR: 30.75 -> 32.91 | Improved |
| add three-phase degradation-aware training (ABAIR) | all-in-one methods · Three-degradation average | PSNR: 32.91 -> 33.21 | Improved |
| use all-in-one model instead of specialized single-task model for dehazing | DehazeFormer vs ABAIR · SOTS dehazing | PSNR: 29.58 -> 33.53 | Improved |
| add CLIP-style foundation-model assistance (DFPIR) versus without | ABAIR · Three-degradation average | PSNR: 33.21 -> 32.88 | No clear effect |
| add vision-language assistance (DA-CLIP + IR-SDE) | all-in-one · Three-degradation average | PSNR: 32.88 -> 27.50 | Hurt |
| use cross-domain dynamic decomposition (D3Net) instead of prompts (PromptIR) | PromptIR · Five-degradation average | PSNR: 29.15 -> 32.36 | Improved |
| use ABAIR over PromptIR in five-degradation setting | PromptIR · Five-degradation average | PSNR: 29.15 -> 31.24 | Improved |
| increase degradations from three to five (Restormer general model) | Restormer · average PSNR across settings | PSNR: 30.75 -> 27.60 | Hurt |
| use all-in-one MODEM instead of early All-in-One model | All-in-One · All-Weather average | PSNR: 28.05 -> 32.87 | Improved |
| use all-in-one LoRA-IR (mixture of low-rank experts) instead of Restormer | Restormer · All-Weather average | PSNR: 29.93 -> 32.76 | Improved |
| add DiffWeather-style foundation/text assistance (CyclicPrompt) | MODEM · All-Weather average | PSNR: 32.87 -> 32.51 | No clear effect |
Closest papers by meaning
- Restore, Assess, Repeat: A Unified Framework for Iterative Image Restoration (CVPR 2026 · similarity 0.93)
- AdaIR: Adaptive All-in-One Image Restoration via Frequency Mining and Modulation (ICLR 2025 · similarity 0.90)
- UARE: A Unified Vision-Language Model for Image Quality Assessment, Restoration, and Enhancement (CVPR 2026 · similarity 0.89)
- Visual-Instructed Degradation Diffusion for All-in-One Image Restoration (CVPR 2025 · similarity 0.89)
- DreamClear: High-Capacity Real-World Image Restoration with Privacy-Safe Dataset Curation (NeurIPS 2024 · similarity 0.88)
- GenDeg: Diffusion-based Degradation Synthesis for Generalizable All-In-One Image Restoration (CVPR 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.
