Research map / Detection and segmentation / CNN features
Dual-Pixel Raindrop Removal
Yizhou Li, Yusuke Monno, Masatoshi Okutomi
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 DP raindrop detection (mask guidance) to the fused removal network (ablation) | DPRRN · Synthetic raindrop (DP, test) | PSNR: 40.93 -> 42.70 | Improved |
| add DP raindrop detection (mask guidance) to the fused removal network (ablation) | DPRRN · Synthetic raindrop (DP, test) | SSIM: 0.9824 -> 0.9867 | Improved |
| add DP raindrop detection (mask guidance) to the fused removal network (ablation) | DPRRN · Real-world raindrop (DP) | PSNR: 31.85 -> 32.47 | Improved |
| add DP raindrop detection (mask guidance) to the fused removal network (ablation) | DPRRN · Real-world raindrop (DP) | SSIM: 0.9357 -> 0.9396 | Improved |
| use dual-pixel left/right inputs instead of single combined image (DPRRN vs best single-image method) | DPRRN vs Restormer · Synthetic raindrop (DP, test) | PSNR: improvements of 0.76dB and 0.0022 | Improved |
| use dual-pixel left/right inputs instead of single combined image (DPRRN vs best single-image method) | DPRRN vs best single-image method · Real-world raindrop (DP) | PSNR: margins of 1.29dB and 0.0137 | Improved |
| use the proposed DPRRN design instead of the prior DP method RDPD+ | DPRRN vs RDPD+ · Synthetic raindrop (DP, test) | PSNR: margins of 2.19dB and 0.0043 | Improved |
| DP fused removal without detection versus single-image Restormer on synthetic data | DPRRN without detection · Synthetic raindrop (DP, test) | PSNR: 41.94 -> 40.93 | Hurt |
| DP fused removal without detection versus single-image Restormer on real-world data | DPRRN without detection · Real-world raindrop (DP) | PSNR: 30.88 -> 31.85 | Improved |
Closest papers by meaning
- Raindrop Clarity: A Dual-Focused Dataset for Day and Night Raindrop Removal (ECCV 2024 · similarity 0.90)
- Unified Removal of Raindrops and Reflections: A New Benchmark and A Novel Pipeline (ECCV 2026 · similarity 0.88)
- Removing Raindrops and Rain Streaks in One Go (CVPR 2021 · similarity 0.88)
- Semantic Segmentation on Raindrop Degraded Images Using Two-Stage Dual Teacher-Student Learning (AAAI 2025 · similarity 0.88)
- Robust Video Content Alignment and Compensation for Rain Removal in a CNN Framework (CVPR 2018 · similarity 0.87)
- Towards Universal Rainy Image Restoration: Benchmark and Baseline (AAAI 2025 · 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.
