Research map / Video / Video segmentation and tracking
Real-Time Intermediate Flow Estimation for Video Frame Interpolation
Zhewei Huang, Tianyuan Zhang, Wen Heng, Boxin Shi, Shuchang Zhou
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 |
|---|---|---|---|
| use privileged teacher distillation instead of self-consistency (ablation) | RIFE · Vimeo90K | PSNR: 35.37 -> 35.61 | Improved |
| use privileged teacher distillation instead of RAFT-KD (ablation) | RIFE · Vimeo90K | PSNR: 35.52 -> 35.61 | Improved |
| use IFNet instead of RAFT + flow reversal layer (ablation) | RIFE · Vimeo90K | PSNR: 35.16 -> 35.61 | Improved |
| use IFNet instead of PWC-Net + forward warping (ablation) | RIFE · Vimeo90K | PSNR: 35.48 -> 35.61 | Improved |
| use three IFBlocks instead of one (ablation) | RIFE · Vimeo90K | PSNR: 35.17 -> 35.61 | Improved |
| use three IFBlocks instead of two (ablation) | RIFE · Vimeo90K | PSNR: 35.46 -> 35.61 | Improved |
| add batch normalization (ablation) | RIFE · Vimeo90K | PSNR: 35.61 -> 35.49 | Hurt |
| use Laplacian loss instead of L1 (ablation) | RIFE · Vimeo90K | PSNR: 35.51 -> 35.61 | Improved |
| use Laplacian loss instead of L1 (ablation) | RIFE · MiddleBury | IE: 1.94 -> 1.96 | No clear effect |
| add fusion mask supervision (ablation) | RIFE · Vimeo90K | PSNR | No clear effect |
| scale loss weights by 10x (ablation) | RIFE · Vimeo90K | convergence | Hurt |
| scale up with test-time augmentation and higher resolution (RIFE-Large) | RIFE · Vimeo90K | PSNR: 35.61 -> 36.19 | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 2 of 2 on most shared results.
- CDFI — CDFI: Compression-Driven Network Design for Frame Interpolation (CVPR 2021) · better on 1 of 1 shared result
- BMBC — BMBC: Bilateral Motion Estimation with Bilateral Cost Volume for Video Interpolation (ECCV 2020) · better on 1 of 1 shared result
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
- Super SloMo: High Quality Estimation of Multiple Intermediate Frames for Video Interpolation (CVPR 2018 · similarity 0.88)
- XVFI: eXtreme Video Frame Interpolation (ICCV 2021 · similarity 0.88)
- BiM-VFI: Bidirectional Motion Field-Guided Frame Interpolation for Video with Non-uniform Motions (CVPR 2025 · similarity 0.87)
- Sparse Global Matching for Video Frame Interpolation with Large Motion (CVPR 2024 · similarity 0.87)
- Disentangled Motion Modeling for Video Frame Interpolation (AAAI 2025 · similarity 0.86)
- IQ-VFI: Implicit Quadratic Motion Estimation for Video Frame Interpolation (CVPR 2024 · similarity 0.86)
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.
