Research map / Adversarial robustness and security / Attacks on text and graphs
Towards Robust Rain Removal Against Adversarial Attacks: A Comprehensive Benchmark Analysis and Beyond
Yi Yu, Wenhan Yang, Yap-Peng Tan, Alex C. Kot
Check 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 |
|---|---|---|---|
| Integrate robust modules and adversarial loss | MPRNet · RainCityscape, LMSE attack | mIoU: 0.3945 -> 0.7592 | Improved |
| Add adversarial loss (ablation) | RCDNet · Rain100H, LMSE attack | PSNR: 12.29 -> 21.85 | Improved |
| Add adversarial loss (ablation) | RCDNet · Rain100H, clean input | SSIM: 0.8867 -> 0.8584 | Hurt |
| Add attention, specifically SE with addition (ablation) | RCDNet · RainCityscape, LMSE attack | mIoU: 0.1924 -> 0.3502 | Improved |
| Add attention, specifically SE with addition (ablation) | RCDNet · RainCityscape, LPIPS attack | AP: 0.5427 -> 0.6730 | Mixed |
| Remove rain-mask side information (ablation) | JORDER-E · RainCityscape, LMSE attack | PSNR: 12.89 -> 14.89 | Improved |
| Remove rain-mask side information (ablation) | JORDER-E · RainCityscape, LMSE attack | AP: 0.6511 -> 0.6202 | Hurt |
| Keep the larger receptive field with dilation paths {1,2,3} (ablation) | JORDER-E · Rain100H, LMSE attack | PSNR: 9.32 -> 10.91 | Improved |
| Add three recurrent blocks (ablation) | Ours · RainCityscape, LMSE attack | mIoU: 0.6512 -> 0.7592 | Improved |
| Add diverse dilations (ablation) | Ours · RainCityscape, LMSE attack | mIoU: 0.7011 -> 0.7592 | Improved |
| Add attention module (ablation) | Ours · RainCityscape, LMSE attack | mIoU: 0.3062 -> 0.7592 | Improved |
| Add adversarial loss (ablation) | Ours · RainCityscape, LMSE attack | mIoU: 0.4745 -> 0.7592 | Improved |
Closest papers by meaning
- Recurrent Multi-Frame Deraining: Combining Physics Guidance and Adversarial Learning (TPAMI 2021 · similarity 0.92)
- Singe Image Rain Removal with Unpaired Information: A Differentiable Programming Perspective (AAAI 2019 · similarity 0.91)
- Deep Joint Rain Detection and Removal From a Single Image (CVPR 2017 · similarity 0.90)
- Joint Rain Detection and Removal from a Single Image with Contextualized Deep Networks (TPAMI 2019 · similarity 0.90)
- Heavy Rain Image Restoration: Integrating Physics Model and Conditional Adversarial Learning (CVPR 2019 · similarity 0.90)
- Learning to Remove Rain in Video With Self-Supervision (TPAMI 2022 · 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.
