Research map / Image generation / Image restoration
Towards Unified Deep Image Deraining: A Survey and a New Benchmark
Xiang Chen, Jinshan Pan, Jiangxin Dong, Jinhui Tang
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 |
|---|---|---|---|
| train on Rain13K instead of no training (zero-shot) (ablation) | Restormer · RainDS-Syn-RS/Real-RS (average) | Avg PSNR: 23.02 -> 27.78 | Improved |
| train on Rain13K instead of Rain200H (ablation) | Restormer · RainDS-Syn-RS/Real-RS (average) | Avg PSNR: 25.00 -> 27.78 | Improved |
| train on Rain200H instead of Rain100H (ablation) | Restormer · RainDS-Syn-RS/Real-RS (average) | Avg PSNR: 24.90 -> 25.00 | Improved |
| train on Rain200L instead of Rain100L (ablation) | Restormer · RainDS-Syn-RS/Real-RS (average) | Avg PSNR: 23.33 -> 23.40 | Improved |
| train on SPA-Data instead of no training (ablation) | Restormer · RainDS-Syn-RS/Real-RS (average) | Avg PSNR: 23.02 -> 23.29 | Improved |
| train on GT-RAIN instead of no training (ablation) | Restormer · RainDS-Syn-RS/Real-RS (average) | Avg PSNR: 23.02 -> 22.80 | Hurt |
| train on DDN-Data instead of Rain13K, real streak test only (ablation) | Restormer · RainDS-Real-RS | PSNR: 24.73 -> 24.74 | Improved |
| train on Rain13K instead of DDN-Data, synthetic test (ablation) | Restormer · RainDS-Syn-RS | PSNR: 29.02 -> 30.83 | Improved |
| use Restormer instead of MAXIM-2S (Transformer vs Transformer) | Mixed track · Rain13K mixed track (average) | Avg PSNR: 33.24 -> 33.96 | Improved |
| use Restormer instead of MPRNet | Mixed track · Rain13K mixed track (average) | Avg PSNR: 32.73 -> 33.96 | Improved |
| use DGUNet+ instead of DGUNet | Mixed track · Rain13K mixed track (average) | Avg PSNR: 33.06 -> 33.46 | Improved |
| use Restormer instead of PReNet | Mixed track · Rain13K mixed track (average) | Avg PSNR: 29.42 -> 33.96 | Improved |
Closest papers by meaning
- Single Image Deraining: From Model-Based to Data-Driven and Beyond (TPAMI 2020 · similarity 0.91)
- Single Image Deraining: A Comprehensive Benchmark Analysis (CVPR 2019 · similarity 0.90)
- Unsupervised Deraining: Where Contrastive Learning Meets Self-Similarity (CVPR 2022 · similarity 0.89)
- Channel Consistency Prior and Self-Reconstruction Strategy Based Unsupervised Image Deraining (CVPR 2025 · similarity 0.89)
- Dreaming To Prune Image Deraining Networks (CVPR 2022 · similarity 0.89)
- Rethinking Multi-Scale Representations in Deep Deraining Transformer (AAAI 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.
