Research map / Video / Video generation
TLB-VFI: Temporal-Aware Latent Brownian Bridge Diffusion for Video Frame Interpolation
Zonglin Lyu, Chen Chen
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 3D wavelet feature gating (ablation) | TLB-VFI · SNU-FILM extreme | FID: 30.717 -> 29.868 | Improved |
| add 3D wavelet feature gating (ablation) | TLB-VFI · Xiph-4K | FID: 19.247 -> 19.114 | Improved |
| add cross-attention aggregation instead of average pooling (ablation) | TLB-VFI · SNU-FILM extreme | FID: 30.903 -> 30.717 | Improved |
| add cross-attention aggregation instead of average pooling (ablation) | TLB-VFI · Xiph-4K | FID: 19.663 -> 19.247 | Improved |
| add temporal attention (ablation) | TLB-VFI · SNU-FILM extreme | FID: 32.061 -> 30.903 | Improved |
| add temporal attention (ablation) | TLB-VFI · Xiph-2K | FID: 10.911 -> 10.499 | Improved |
| use 3D convolution instead of 2D (ablation) | TLB-VFI · Xiph-4K | FID: 23.481 -> 19.944 | Improved |
| use 3D convolution instead of 2D (ablation) | TLB-VFI · SNU-FILM extreme | FID: 33.155 -> 32.061 | Improved |
| make the image encoder temporal-aware (zero-replaced frame corrupts multi-level features) (ablation) | TLB-VFI · Xiph-4K | FID: 19.114 -> 22.731 | Hurt |
| make the image encoder temporal-aware (ablation) | TLB-VFI · SNU-FILM extreme | FID: 29.868 -> 34.982 | Hurt |
| enable feature sharing in the temporal-aware image encoder variant (ablation) | TLB-VFI · Xiph-4K | FID: 20.004 -> 22.731 | Hurt |
| replace Consec. BB adjacent-frame bridge with zero-replaced frame bridge and temporal autoencoder (full TLB-VFI) | Consec. BB · SNU-FILM extreme | FID: 36.631 -> 29.868 | Improved |
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.
- LDMVFI — LDMVFI: Video Frame Interpolation with Latent Diffusion Models (AAAI 2024) · 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
- EDEN: Enhanced Diffusion for High-quality Large-motion Video Frame Interpolation (CVPR 2025 · similarity 0.91)
- LDMVFI: Video Frame Interpolation with Latent Diffusion Models (AAAI 2024 · similarity 0.91)
- TRIP: Temporal Residual Learning with Image Noise Prior for Image-to-Video Diffusion Models (CVPR 2024 · similarity 0.90)
- Efficient Video Diffusion Models via Content-Frame Motion-Latent Decomposition (ICLR 2024 · similarity 0.90)
- Hierarchical Flow Diffusion for Efficient Frame Interpolation (CVPR 2025 · similarity 0.90)
- Repurposing Pre-trained Video Diffusion Models for Event-based Video Interpolation (CVPR 2025 · similarity 0.89)
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.
