Research map / Video / Video segmentation and tracking
Bringing Events Into Video Deblurring With Non-Consecutively Blurry Frames
Wei Shang, Dongwei Ren, Dongqing Zou, Jimmy S. Ren, Ping Luo, 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 FTCE (temporal consistency enhancement) to FDET+FBRN (ablation) | D2Nets · GoPro (whole videos) | PSNR: 31.00 -> 31.60 | Improved |
| add FDET and FBRN in front of FTCE (ablation) | D2Nets · GoPro (whole videos) | PSNR: 31.07 -> 31.60 | Improved |
| add EFM events fusion to D2Nets (ablation) | D2Nets · GoPro (whole videos) | PSNR: 31.60 -> 31.76 | Improved |
| use EFM instead of naive concatenation of events and frames (ablation) | D2Nets* · GoPro | PSNR: -0.42dB PSNR decrease when replaced by concatenation | Improved |
| feed events instead of frames to BiLSTM detector | BiLSTM detector · GoPro | detection accuracy: 97.2% -> 99.0% | Improved |
| feed events instead of frames to BiLSTM detector | BiLSTM detector · Blur-DVS | detection accuracy: 94.8% -> 97.6% | Improved |
| use BiLSTM instead of LSTM in the detector (ablation) | blur detector · GoPro | detection accuracy: 2.98% decrease for frames as input and 3.84% decrease for events as input with LSTM | Improved |
| add EFM to DMPHN | DMPHN · GoPro (whole videos) | PSNR: 31.58 -> 31.90 | Improved |
| add EFM to STFAN | STFAN · GoPro (whole videos) | PSNR: 30.12 -> 30.90 | Improved |
| add EFM to CDVD-TSP | CDVD-TSP · GoPro (whole videos) | PSNR: 30.31 -> 32.24 | Improved |
| add EFM to D2Nets | D2Nets · GoPro (blurry frames only) | PSNR / SSIM: 27.68 -> 27.39 PSNR; 0.906 -> 0.907 SSIM | Mixed |
| add EFM to DMPHN | DMPHN · Blur-DVS (whole videos) | PSNR: 29.10 -> 31.36 | Improved |
Closest papers by meaning
- Event-Based Fusion for Motion Deblurring with Cross-Modal Attention (ECCV 2022 · similarity 0.92)
- Frequency-aware Event-based Video Deblurring for Real-World Motion Blur (CVPR 2024 · similarity 0.91)
- Generalizing Event-Based Motion Deblurring in Real-World Scenarios (ICCV 2023 · similarity 0.90)
- Diffusion-based Event Generation for High-Quality Image Deblurring (CVPR 2025 · similarity 0.89)
- Event-Enhanced Blurry Video Super-Resolution (AAAI 2025 · similarity 0.88)
- Motion Deblurring via Spatial-Temporal Collaboration of Frames and Events (AAAI 2024 · similarity 0.88)
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.
