Saved in:
| Main Authors: | , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2503.15831 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866910935448289280 |
|---|---|
| author | Zhang, Zihao Chen, Haoran Zhao, Haoyu Lu, Guansong Fu, Yanwei Xu, Hang Wu, Zuxuan |
| author_facet | Zhang, Zihao Chen, Haoran Zhao, Haoyu Lu, Guansong Fu, Yanwei Xu, Hang Wu, Zuxuan |
| contents | Handling complex or nonlinear motion patterns has long posed challenges for video frame interpolation. Although recent advances in diffusion-based methods offer improvements over traditional optical flow-based approaches, they still struggle to generate sharp, temporally consistent frames in scenarios with large motion. To address this limitation, we introduce EDEN, an Enhanced Diffusion for high-quality large-motion vidEo frame iNterpolation. Our approach first utilizes a transformer-based tokenizer to produce refined latent representations of the intermediate frames for diffusion models. We then enhance the diffusion transformer with temporal attention across the process and incorporate a start-end frame difference embedding to guide the generation of dynamic motion. Extensive experiments demonstrate that EDEN achieves state-of-the-art results across popular benchmarks, including nearly a 10% LPIPS reduction on DAVIS and SNU-FILM, and an 8% improvement on DAIN-HD. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2503_15831 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | EDEN: Enhanced Diffusion for High-quality Large-motion Video Frame Interpolation Zhang, Zihao Chen, Haoran Zhao, Haoyu Lu, Guansong Fu, Yanwei Xu, Hang Wu, Zuxuan Computer Vision and Pattern Recognition Handling complex or nonlinear motion patterns has long posed challenges for video frame interpolation. Although recent advances in diffusion-based methods offer improvements over traditional optical flow-based approaches, they still struggle to generate sharp, temporally consistent frames in scenarios with large motion. To address this limitation, we introduce EDEN, an Enhanced Diffusion for high-quality large-motion vidEo frame iNterpolation. Our approach first utilizes a transformer-based tokenizer to produce refined latent representations of the intermediate frames for diffusion models. We then enhance the diffusion transformer with temporal attention across the process and incorporate a start-end frame difference embedding to guide the generation of dynamic motion. Extensive experiments demonstrate that EDEN achieves state-of-the-art results across popular benchmarks, including nearly a 10% LPIPS reduction on DAVIS and SNU-FILM, and an 8% improvement on DAIN-HD. |
| title | EDEN: Enhanced Diffusion for High-quality Large-motion Video Frame Interpolation |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2503.15831 |