Research map / Detection and segmentation / CNN features
Multi-Scale Progressive Fusion Network for Single Image Deraining
Check your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 multi-scale progressive fusion instead of the comparison fusion strategy (ablation) | MSPFN · Test100 | PSNR: +0.54dB | Helped |
| add CFM (ablation) | MSPFN · Test100 | PSNR: 27.01 -> 27.29 | Helped |
| add FFMs (ablation) | MSPFN · Test100 | PSNR: 23.69 -> 27.29 | Helped |
| use a multi-scale pyramid instead of a single-scale original-image input (ablation) | MSPFN · Test100 | PSNR: 26.56 -> 27.29 | Helped |
| remove edge constraint from the loss (ablation) | MSPFN · Test100 | PSNR: 26.93 -> 27.50 | Hurt |
| increase FFM depth from M=10 to M=30 (N=1) (ablation) | MSPFN · Test100 | PSNR: 27.29 -> 27.91 | Helped |
| increase CAU count from N=1 to N=2 at M=17 (ablation) | MSPFN · Test100 | PSNR: +0.13dB | Helped |
| reduce FFM depth from M=17 to M=13 while using N=2 (ablation) | MSPFN · Test100 | PSNR: 27.63 -> 27.42 | Hurt |
| use the lightweight M5N1 variant with narrower filters (ablation) | MSPFN · Test100 | PSNR: 27.29 -> 24.99 | Hurt |
| use MSPFN instead of PreNet | single-image deraining · Rain100H | PSNR: 26.77 -> 28.66 | Helped |
| use MSPFN instead of PreNet | single-image deraining · Rain100L | PSNR: 32.44 -> 32.40 | Hurt |
| use MSPFN instead of PreNet | single-image deraining · Test2800 | PSNR: 31.75 -> 32.82 | Helped |
Closest papers by meaning
- Hybrid CNN-Transformer Feature Fusion for Single Image Deraining (AAAI 2023 · similarity 0.92)
- Spatial Attentive Single-Image Deraining With a High Quality Real Rain Dataset (CVPR 2019 · similarity 0.92)
- Deep Joint Rain Detection and Removal From a Single Image (CVPR 2017 · similarity 0.91)
- Bidirectional Multi-Scale Implicit Neural Representations for Image Deraining (CVPR 2024 · similarity 0.91)
- Boosting Image De-Raining via Central-Surrounding Synergistic Convolution (AAAI 2025 · similarity 0.90)
- Harnessing Joint Rain-/Detail-aware Representations to Eliminate Intricate Rains (ICLR 2024 · similarity 0.90)
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.