Research map / Video / Video segmentation and tracking
SelfHVD: Self-Supervised Handheld Video Deblurring
Honglei Xu, Zhilu Zhang, Junjie Fan, Xiaohe Wu, Wangmeng Zuo
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 SEVD (ablation) | SelfHVD BasicVSR++ · GoProShake | PSNR: 35.61 -> 36.67 | Improved |
| add SCSCM (ablation) | SelfHVD BasicVSR++ · GoProShake | PSNR: 35.61 -> 37.09 | Improved |
| add SCSCM on top of SEVD (ablation) | SelfHVD BasicVSR++ · GoProShake | PSNR: 36.67 -> 37.44 | Improved |
| add RSCR (random sharp clue removal) (ablation) | SelfHVD BasicVSR++ with SEVD · GoProShake | PSNR: 35.11 -> 37.44 | Improved |
| add SIS supervision selection (ablation) | SelfHVD BasicVSR++ with RSCR · GoProShake | PSNR: 36.27 -> 37.44 | Improved |
| increase SCSCM updates from 1 to 2 (ablation) | SelfHVD BasicVSR++ · GoProShake | PSNR: 36.00 -> 37.44 | Improved |
| increase SCSCM updates from 2 to 3 (ablation) | SelfHVD BasicVSR++ · GoProShake | PSNR: 37.44 -> 37.57 | No clear effect |
| add uncertainty and occlusion masks (ablation) | SelfHVD BasicVSR++ · GoProShake | PSNR: 36.13 -> 37.44 | Improved |
| use SEA-RAFT instead of RAFT for alignment (ablation) | SelfHVD BasicVSR++ · GoProShake | PSNR: 36.78 -> 37.44 | Improved |
| use SEA-RAFT instead of FlowFormer++ (ablation) | SelfHVD BasicVSR++ · GoProShake | PSNR: 37.24 -> 37.44 | Improved |
| sharp frame selection interval 20 instead of 30 (ablation) | SelfHVD BasicVSR++ · GoProShake | PSNR: 37.03 -> 37.44 | Improved |
| sharp frame selection interval 20 instead of 5 (ablation) | SelfHVD BasicVSR++ · GoProShake | PSNR: 37.33 -> 37.44 | Improved |
Closest papers by meaning
- Motion Deblurring With Real Events (ICCV 2021 · similarity 0.87)
- Learning Blind Motion Deblurring (ICCV 2017 · similarity 0.87)
- EVDI++: Event-Based Video Deblurring and Interpolation via Self-Supervised Learning (TPAMI 2026 · similarity 0.87)
- ARVo: Learning All-Range Volumetric Correspondence for Video Deblurring (CVPR 2021 · similarity 0.87)
- Flow-Guided Sparse Transformer for Video Deblurring (ICML 2022 · similarity 0.86)
- Frequency-aware Event-based Video Deblurring for Real-World Motion Blur (CVPR 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.
