Research map / Video / Video segmentation and tracking
Progressive Temporal Feature Alignment Network for Video Inpainting
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 |
|---|---|---|---|
| add optical-flow alignment and a validity mask to TSM (ablation) | Progressive Temporal Feature Alignment Network · FVI, Object Mask | PSNR: 35.06 -> 35.48 | Helped |
| add optical-flow alignment and a validity mask to TSM (ablation) | Progressive Temporal Feature Alignment Network · FVI, Curve Mask | PSNR: 37.12 -> 37.43 | Helped |
| add optical-flow alignment and a validity mask to TSM (ablation) | Progressive Temporal Feature Alignment Network · FVI, Stationary Mask | PSNR: 41.11 -> 41.41 | Helped |
| replace completed optical flow with ground-truth optical flow (ablation) | Progressive Temporal Feature Alignment Network · FVI, Object Mask | PSNR: 35.48 -> 35.76 | Helped |
| replace completed optical flow with ground-truth optical flow (ablation) | Progressive Temporal Feature Alignment Network · FVI, Curve Mask | PSNR: 37.43 -> 37.82 | Helped |
| replace completed optical flow with ground-truth optical flow (ablation) | Progressive Temporal Feature Alignment Network · FVI, Stationary Mask | PSNR: 41.41 -> 41.61 | Helped |
| use completed flow rather than ground-truth optical flow (ablation) | Progressive Temporal Feature Alignment Network · FVI, Object Mask | PSNR: 35.76 -> 35.48 | Hurt |
| use completed flow rather than ground-truth optical flow (ablation) | Progressive Temporal Feature Alignment Network · FVI, Curve Mask | PSNR: 37.82 -> 37.43 | Hurt |
| use completed flow rather than ground-truth optical flow (ablation) | Progressive Temporal Feature Alignment Network · FVI, Stationary Mask | PSNR: 41.61 -> 41.41 | Hurt |
Closest papers by meaning
- Video Inpainting by Jointly Learning Temporal Structure and Spatial Details (AAAI 2019 · similarity 0.90)
- Deep Video Inpainting (CVPR 2019 · similarity 0.87)
- Deep Stereo Video Inpainting (CVPR 2023 · similarity 0.87)
- LAMV: Learning to Align and Match Videos With Kernelized Temporal Layers (CVPR 2018 · similarity 0.86)
- Inertia-Guided Flow Completion and Style Fusion for Video Inpainting (CVPR 2022 · similarity 0.86)
- Deep Flow-Guided Video Inpainting (CVPR 2019 · 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.