Research map / Video / Video generation
Towards Holistic Modeling for Video Frame Interpolation with Auto-regressive Diffusion Transformers
Xinyu Peng, Han Li, Yuyang Huang, Ziyang Zheng, Yaoming Wang, Xin Chen, Wenrui Dai, Chenglin Li, Junni Zou, Hongkai Xiong
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 skip-concatenate sampling instead of causal AR order (ablation) | LDF-VFI · SNU-FILM-entire (average) | FVD: 22.67 -> 17.05 | Improved |
| add skip-concatenate sampling instead of causal AR order (ablation) | LDF-VFI · SNU-FILM-entire (average) | LPIPS: 0.062 -> 0.060 | Improved |
| add conditional VAE decoder over Wan2.1 decoder (ablation) | LDF-VFI · SNU-FILM-entire (average) | LPIPS: 0.060 -> 0.051 | Improved |
| add conditional VAE decoder over Wan2.1 decoder (ablation) | LDF-VFI · SNU-FILM-entire (average) | FVD: 17.05 -> 16.49 | Improved |
| add conditional VAE decoder over Wan2.1 decoder (ablation) | LDF-VFI · SNU-FILM-entire 8x | FVD: 14.55 -> 15.01 | Hurt |
| use sparse attention instead of full attention (ablation) | LDF-VFI · X4K-entire 16x | FVD: 772.32 -> 120.40 | Improved |
| use sparse attention instead of full attention (ablation) | LDF-VFI · SNU-FILM-entire 8x | running time (s): 4.68 -> 3.36 | Improved |
| use nearest-neighbor upsampling instead of zero padding (ablation) | LDF-VFI · SNU-FILM-4x key-frame VAE reconstruction | PSNR: 28.40 -> 30.87 | Improved |
| use temporal chunk length 20 instead of 5 (ablation) | LDF-VFI · X4K-entire 16x | FVD: 510.63 -> 121.29 | Improved |
| use temporal chunk length 20 instead of 5 (ablation) | LDF-VFI · SNU-FILM-entire 8x | LPIPS: 0.059 -> 0.066 | Hurt |
| increase sampling steps from 2 to 16 | LDF-VFI · SNU-FILM-entire 8x | FVD: 21.25 -> 15.01 | 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.
- BiM-VFI — BiM-VFI: Bidirectional Motion Field-Guided Frame Interpolation for Video with Non-uniform Motions (CVPR 2025) · better on 3 of 4 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
- LDMVFI: Video Frame Interpolation with Latent Diffusion Models (AAAI 2024 · similarity 0.91)
- Repurposing Pre-trained Video Diffusion Models for Event-based Video Interpolation (CVPR 2025 · similarity 0.91)
- EDEN: Enhanced Diffusion for High-quality Large-motion Video Frame Interpolation (CVPR 2025 · similarity 0.91)
- Disentangled Motion Modeling for Video Frame Interpolation (AAAI 2025 · similarity 0.89)
- Generative Inbetweening: Adapting Image-to-Video Models for Keyframe Interpolation (ICLR 2025 · similarity 0.89)
- ViBiDSampler: Enhancing Video Interpolation Using Bidirectional Diffusion Sampler (ICLR 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.
