Research map / Video / Video segmentation and tracking
One-Shot Flow, Any-Time Frame: A Bidirectional Warping Framework for Event-Based Video Frame Interpolation
Linghui Fu, Yuhan Liu, Hao Chen, Zhen Yang, Yongjian Deng
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 |
|---|---|---|---|
| replace BiFEB with RAFT forward flow + Timelens backward flow (ablation) | Ours · GOPRO | PSNR: 36.96 -> 36.27 | Hurt |
| use BiFEB with n=4 instead of n=16 (ablation) | Ours · GOPRO | PSNR: 36.96 -> 29.75 | Hurt |
| use BiFEB with n=8 instead of n=16 (ablation) | Ours · GOPRO | PSNR: 36.96 -> 35.24 | Hurt |
| use bidirectional warping instead of forward warping only (ablation) | Ours · GOPRO | PSNR: 30.54 -> 36.96 | Improved |
| use bidirectional warping instead of backward warping only (ablation) | Ours · GOPRO | PSNR: 35.78 -> 36.96 | Improved |
| add hole mask Rh in BiW (ablation) | Ours · GOPRO | PSNR: 36.01 -> 36.96 | Improved |
| add difference mask Rd in BiW (ablation) | Ours · GOPRO | PSNR: 36.74 -> 36.96 | Improved |
| one-shot flow query vs per-frame flow estimation as frame count grows | Ours · GOPRO | time per interpolated frame: 0.137s -> 0.108s | Improved |
| use Ours instead of TLXNet when interpolating 127 frames | Ours · GOPRO 127 frames | memory: TLXNet OOM vs 7.27GB | Improved |
| use Ours instead of TimeTracker on BS-ERGB Skip 1 | Ours · BS-ERGB Skip 1 | PSNR: 29.85 -> 29.76 | Hurt |
| use Ours instead of TimeTracker on HS-ERGB Skip 7 | Ours · HS-ERGB Skip 7 | PSNR: 32.68 -> 34.19 | Improved |
| use Ours instead of TLXNet at 31 frames | Ours · GOPRO 31 frames | time per interpolated frame: 0.079s -> 0.137s | Hurt |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- TimeTracker — TimeTracker: Event-based Continuous Point Tracking for Video Frame Interpolation with Non-linear Motion (CVPR 2025) · better on 6 of 8 shared results
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- Video Frame Interpolation via Direct Synthesis with the Event-based Reference (CVPR 2024 · similarity 0.90)
- Time Lens++: Event-Based Frame Interpolation With Parametric Non-Linear Flow and Multi-Scale Fusion (CVPR 2022 · similarity 0.90)
- Repurposing Pre-trained Video Diffusion Models for Event-based Video Interpolation (CVPR 2025 · similarity 0.89)
- TimeTracker: Event-based Continuous Point Tracking for Video Frame Interpolation with Non-linear Motion (CVPR 2025 · similarity 0.88)
- BiM-VFI: Bidirectional Motion Field-Guided Frame Interpolation for Video with Non-uniform Motions (CVPR 2025 · similarity 0.88)
- Video Frame Prediction from a Single Image and Events (AAAI 2024 · similarity 0.88)
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.
