Research map / Video / Video segmentation and tracking
Context-Aware Video Reconstruction for Rolling Shutter 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 context-aware reconstruction with motion enhancement and occlusion reasoning instead of RSSR | Rolling-shutter video reconstruction · Carla-RS with occlusion mask (CRM), t=0.5 | PSNR: 30.17 -> 32.02 | Helped |
| Use context-aware reconstruction with motion enhancement and occlusion reasoning instead of SUNet | Rolling-shutter video reconstruction · Carla-RS without occlusion mask (CR), t=0.5 | PSNR: 29.18 -> 31.74 | Helped |
| Use context-aware reconstruction with motion enhancement and occlusion reasoning instead of SUNet | Rolling-shutter video reconstruction · Fastec-RS (FR), t=0.5 | PSNR: 28.34 -> 28.72 | Helped |
| Add a learned NBMF initialization rather than use ABMF | CVR · Carla-RS with occlusion mask (CRM), t=0.5 | PSNR: 31.82 -> 32.02 | Helped |
| Replace the NBMF/ABMF initialization with linear BMF (ablation) | CVR · Carla-RS with occlusion mask (CRM), t=0.5 | PSNR: 32.02 -> 26.10 | Hurt |
| Replace PWC-Net with RAFT (ablation) | CVR · Carla-RS with occlusion mask (CRM), t=0.5 | PSNR: 32.02 -> 30.50 | Hurt |
| Train the motion interpretation module jointly instead of freezing it (ablation) | CVR · Fastec-RS (FR), t=0.5 | PSNR: 28.11 -> 28.72 | Helped |
| Add motion enhancement residuals ΔU (ablation) | CVR · Fastec-RS (FR), t=0.5 | PSNR: 28.32 -> 28.72 | Helped |
| Add bilateral occlusion masks (ablation) | CVR · Fastec-RS (FR), t=0.5 | PSNR: 24.04 -> 28.72 | Helped |
| Add the self-supervised candidate-consistency loss L'c (ablation) | CVR · Fastec-RS (FR), t=0.5 | PSNR: 28.72 -> 28.55 | No clear effect |
Closest papers by meaning
- Rolling Shutter Correction with Intermediate Distortion Flow Estimation (CVPR 2024 · similarity 0.88)
- Motion Blur Decomposition with Cross-shutter Guidance (CVPR 2024 · similarity 0.88)
- Single Image Rolling Shutter Removal with Diffusion Models (AAAI 2025 · similarity 0.86)
- Order-One Rolling Shutter Cameras (CVPR 2025 · similarity 0.85)
- USB-NeRF: Unrolling Shutter Bundle Adjusted Neural Radiance Fields (ICLR 2024 · similarity 0.84)
- BARD-GS: Blur-Aware Reconstruction of Dynamic Scenes via Gaussian Splatting (CVPR 2025 · similarity 0.84)
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.