Research map / Video / Video segmentation and tracking
Video Frame Interpolation With Transformer
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 VFIformer instead of the strongest listed baseline | VFI on Vimeo90K test · Vimeo90K test | PSNR: 36.18 -> 36.50 | Helped |
| Add Transformer layers in place of convolutional layers (ablation) | VFIformer · Vimeo90K test | PSNR: 36.27 -> 36.49 | Helped |
| Add cross-scale window attention instead of standard window attention (ablation) | VFIformer · SNU-FILM Easy | PSNR: 40.06 -> 40.13; 0.07 dB | Helped |
| Add cross-scale window attention instead of standard window attention (ablation) | VFIformer · SNU-FILM Medium | PSNR: 36.03 -> 36.09; 0.06 dB | Helped |
| Add cross-scale window attention instead of standard window attention (ablation) | VFIformer · SNU-FILM Hard | PSNR: 30.61 -> 30.67; 0.06 dB | Helped |
| Increase attention window size from 4 to 8 (ablation) | VFIformer · Vimeo90K test | PSNR: 36.24 -> 36.29 | Helped |
| Increase attention window size from 8 to 12 (ablation) | VFIformer · Vimeo90K test | PSNR: 36.29 -> 36.31 | Helped |
| Increase TFL count from 2 to 4 (ablation) | VFIformer · Vimeo90K test | PSNR: 36.35 -> 36.43 | Helped |
| Increase TFL count from 4 to 6 (ablation) | VFIformer · Vimeo90K test | PSNR: 36.43 -> 36.49 | Helped |
| Increase TFL count from 6 to 8 (ablation) | VFIformer · Vimeo90K test | PSNR: 36.49 -> 36.52 | Helped |
| Add CSWA instead of standard window attention (ablation) | VFIformer · Vimeo90K test, flow range [0,3) | PSNR: 36.51 -> 36.52; 0.01 | Helped |
| Add CSWA instead of standard window attention (ablation) | VFIformer · Vimeo90K test, flow range [10,+∞) | PSNR: 34.00 -> 34.04; 0.04 | Helped |
Closest papers by meaning
- LDMVFI: Video Frame Interpolation with Latent Diffusion Models (AAAI 2024 · similarity 0.89)
- Disentangled Motion Modeling for Video Frame Interpolation (AAAI 2025 · similarity 0.88)
- Repurposing Pre-trained Video Diffusion Models for Event-based Video Interpolation (CVPR 2025 · similarity 0.87)
- VFIMamba: Video Frame Interpolation with State Space Models (NeurIPS 2024 · similarity 0.87)
- Perception-Oriented Video Frame Interpolation via Asymmetric Blending (CVPR 2024 · similarity 0.87)
- Sparse Global Matching for Video Frame Interpolation with Large Motion (CVPR 2024 · similarity 0.87)
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.