Research map / Continual learning / Online and replay-based
Improving De-Raining Generalization via Neural Reorganization
Jie Xiao, Man Zhou, Xueyang Fu, Aiping Liu, Zheng-Jun Zha
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 Synaptic Consolidation on top of knowledge distillation (ablation) | PreNet with NR · SPA-Data | PSNR: 35.14 -> 35.60 | Improved |
| add Knowledge Distillation on top of synaptic consolidation (ablation) | PreNet with NR · SPA-Data | PSNR: 35.20 -> 35.60 | Improved |
| add both SC and KD modules to sequential training (ablation) | PreNet · SPA-Data | PSNR: 34.91 -> 35.60 | Improved |
| add both SC and KD modules to sequential training (ablation) | PreNet · SPA-Data | SSIM: 0.951 -> 0.959 | Improved |
| train with NR instead of mixing the four datasets | PreNet · SPA-Data | PSNR: 34.85 -> 35.60 | Improved |
| train with NR instead of mixing the four datasets | PreNet · Rain100L | PSNR: 31.53 -> 36.34 | Improved |
| train with NR instead of mixing the four datasets | PreNet · Rain100H | PSNR: 26.74 -> 25.93 | Hurt |
| train with NR instead of naive sequential training | PreNet · Rain100H | PSNR: 18.01 -> 25.93 | Improved |
| train with NR instead of naive sequential training | PreNet · SPA-Data | PSNR: 34.91 -> 35.60 | Improved |
| add more datasets progressively in NR sequence (Rain100H to Rain100L) (ablation) | PreNet with NR · SPA-Data | PSNR: 34.50 -> 35.60 | Improved |
| train on a single synthetic dataset (Rain100H) vs multi-dataset NR | PreNet · SPA-Data | PSNR: 34.50 -> 35.60 | Improved |
| use NR instead of Syn2Real transfer learning (no real images) | PreNet · SPA-Data | PSNR: 32.87 -> 32.97 | Improved |
Closest papers by meaning
- Image De-Raining via Continual Learning (CVPR 2021 · similarity 0.89)
- Incremental Image De-raining via Associative Memory (AAAI 2023 · similarity 0.88)
- Boosting Image De-Raining via Central-Surrounding Synergistic Convolution (AAAI 2025 · similarity 0.88)
- Harnessing Joint Rain-/Detail-aware Representations to Eliminate Intricate Rains (ICLR 2024 · similarity 0.87)
- Bidirectional Multi-Scale Implicit Neural Representations for Image Deraining (CVPR 2024 · similarity 0.87)
- Neuromorphic Event Signal-Driven Network for Video De-raining (AAAI 2024 · similarity 0.86)
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.
