Research map / Video / Video generation
OMoBlur: An Object Motion Blur Dataset and Benchmark for Real-World Local Motion Deblurring
Dingchuan Yu, Jiatong Li, Jingwen Zhou, Zhengyue Zhuge, Yueting Chen, Qi Li
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 Motion-Appearance Extract Block to backbone (ablation) | OMDNet · OMoBlur test set | weighted PSNR: 33.39 -> 33.46 | Improved |
| add flow-guided gate predictor (ablation) | OMDNet (MAEB) · OMoBlur test set | weighted PSNR: 33.46 -> 33.88 | Improved |
| add Differential Transposed Attention Module (ablation) | OMDNet (MAEB+FGP) · OMoBlur test set | weighted PSNR: 33.88 -> 34.04 | Improved |
| add all modules versus minimalist backbone (ablation) | OMDNet · OMoBlur test set | PSNR: 35.53 -> 35.73 | Improved |
| use adaptive gated fusion instead of LBAG gating (ablation) | OMDNet · ReLoBlur static region (zoom 1) | MAE vs input: 0.023 -> 0.007 | Improved |
| add gating versus no gating (ablation) | OMDNet · ReLoBlur static region (zoom 1) | MAE vs input: 0.041 -> 0.007 | Improved |
| train on OMoBlur instead of GoPro | MIMO-UNet and Restormer · ReLoBlur real blur (qualitative) | visual deblurring | Improved |
| train on physically accumulated OMoBlur instead of frame-interpolated OMoBlur-FI (ablation) | MIMO-UNet and Restormer · ReLoBlur real blur (qualitative) | visual deblurring | Improved |
| OMDNet versus LMD-ViT on global PSNR | OMDNet · OMoBlur test set | PSNR: 35.68 -> 35.73 | No clear effect |
Closest papers by meaning
- Real-World Deep Local Motion Deblurring (AAAI 2023 · similarity 0.91)
- Parameterized Blur Kernel Prior Learning for Local Motion Deblurring (CVPR 2025 · similarity 0.89)
- Adaptive Window Pruning for Efficient Local Motion Deblurring (ICLR 2024 · similarity 0.88)
- Motion-adaptive Separable Collaborative Filters for Blind Motion Deblurring (CVPR 2024 · similarity 0.88)
- Real-World Efficient Blind Motion Deblurring via Blur Pixel Discretization (CVPR 2024 · similarity 0.88)
- DynaMoDe-NeRF: Motion-aware Deblurring Neural Radiance Field for Dynamic Scenes (CVPR 2025 · similarity 0.87)
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.
