Research map / Video / Video segmentation and tracking
Error Compensation Framework for Flow-Guided Video Inpainting
Jaeyeon Kang, Seoung Wug Oh, Seon Joo Kim
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 compensation stage (ablation) | ECFVI · Youtube-VI stationary mask | PSNR: 30.05 -> 33.53 | Improved |
| add compensation stage (ablation) | ECFVI · Youtube-VI moving mask | PSNR: 34.04 -> 36.80 | Improved |
| add error guidance map (ablation) | ECFVI · Youtube-VI stationary mask | PSNR: 30.48 -> 33.53 | Improved |
| add error guidance map (ablation) | ECFVI · Youtube-VI moving mask | PSNR: 33.81 -> 36.80 | Improved |
| add local temporal network to flow completion (ablation) | ECFVI · Youtube-VI stationary mask | Flow-EPE: 0.71 -> 0.33 | Improved |
| add local temporal network to flow completion (ablation) | ECFVI · Youtube-VI moving mask | Flow-EPE: 0.83 -> 0.49 | Improved |
| add hard-region weighting in flow loss (ablation) | ECFVI · Youtube-VI moving mask | Flow-EPE: 0.58 -> 0.49 | Improved |
| add hard-region weighting in flow loss (ablation) | ECFVI · Youtube-VI stationary mask | Flow-EPE: 0.40 -> 0.33 | Improved |
| use jointly trained RGB-guided flow completion instead of FuseFormer + RAFT (ablation) | ECFVI · Youtube-VI moving mask | Flow-EPE: 0.80 -> 0.49 | Improved |
| replace FGVC flow completion by compensation pipeline (ablation) | FGVC* · Youtube-VI stationary mask | PSNR: 31.21 -> 32.87 | Improved |
| swap own flow completion for RAFT pseudo ground-truth flow (ablation) | ECFVI · Youtube-VI stationary mask | PSNR: 33.53 -> 33.64 | Improved |
| swap own flow completion for RAFT pseudo ground-truth flow (ablation) | ECFVI · Youtube-VI moving mask | PSNR: 36.80 -> 37.17 | Improved |
Closest papers by meaning
- Elevating Flow-Guided Video Inpainting with Reference Generation (AAAI 2025 · similarity 0.88)
- FlowVid: Taming Imperfect Optical Flows for Consistent Video-to-Video Synthesis (CVPR 2024 · similarity 0.87)
- Video Diffusion Models Are Strong Video Inpainter (AAAI 2025 · similarity 0.86)
- AVID: Any-Length Video Inpainting with Diffusion Model (CVPR 2024 · similarity 0.85)
- Sparse Global Matching for Video Frame Interpolation with Large Motion (CVPR 2024 · similarity 0.85)
- Keyframe-Guided Creative Video Inpainting (CVPR 2025 · similarity 0.85)
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.
