Research map / Video / Video segmentation and tracking
Deep Video Deblurring for Hand-Held Cameras
Check your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 |
|---|---|---|---|
| use neighboring frames instead of five copies of the central frame (ablation) | DeBlurNet · 10 test datasets, average | PSNR: 28.37 -> 30.05 | Helped |
| use neighboring frames instead of five copies of the central frame (ablation) | DeBlurNet · 10 test datasets, average | MSSIM: .941 -> .964 | Helped |
| add optical-flow alignment to the neighboring-frame inputs (ablation) | DeBlurNet · 10 test datasets, average | MSSIM: .964 -> .969 | Helped |
| add optical-flow alignment to the neighboring-frame inputs (ablation) | DeBlurNet · 10 test datasets, average | PSNR: 30.05 -> 30.05 | No clear effect |
| add global homography alignment to the neighboring-frame inputs (ablation) | DeBlurNet · 10 test datasets, average | PSNR: 30.05 -> 29.73 | Hurt |
| add global homography alignment to the neighboring-frame inputs (ablation) | DeBlurNet · 10 test datasets, average | MSSIM: .964 -> .962 | Hurt |
| use no alignment rather than optical-flow alignment (ablation) | DeBlurNet · input frames with high PSNR, especially above 29dB | PSNR | Helped |
| use optical-flow alignment rather than no alignment (ablation) | DeBlurNet · severely blurred input frames | PSNR | Helped |
| use late fusion of deeper features rather than early input fusion (ablation) | DeBlurNet · test and validation data | PSNR | Hurt |
| add late fusion of deeper features (ablation) | DeBlurNet · challenging cases where DBN+NOALIGN fails | deblurring quality | Mixed |
| add multiscale-supervised Laplacian reconstruction (ablation) | DeBlurNet · video deblurring | deblurring quality | Mixed |
| predict sharp Fourier coefficients instead of output pixels (ablation) | DeBlurNet · video deblurring | deblurring quality | Hurt |
Closest papers by meaning
- Deep Multi-Scale Convolutional Neural Network for Dynamic Scene Deblurring (CVPR 2017 · similarity 0.90)
- Cascaded Deep Video Deblurring Using Temporal Sharpness Prior (CVPR 2020 · similarity 0.89)
- Deep Recurrent Neural Network with Multi-Scale Bi-directional Propagation for Video Deblurring (AAAI 2022 · similarity 0.89)
- ARVo: Learning All-Range Volumetric Correspondence for Video Deblurring (CVPR 2021 · similarity 0.89)
- Efficient Dynamic Scene Deblurring Using Spatially Variant Deconvolution Network With Optical Flow Guided Training (CVPR 2020 · similarity 0.89)
- Parameterized Blur Kernel Prior Learning for Local Motion Deblurring (CVPR 2025 · similarity 0.89)
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.