Uniform Discrete Diffusion with Metric Path for Video Generation
Fuente:
arXiv
Salvato in:
| Autori principali: | , , , , , , , , , , |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2025
|
| Soggetti: | |
| Accesso online: | |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866911237251530752 |
|---|---|
| author | Deng, Haoge Pan, Ting Zhang, Fan Liu, Yang Luo, Zhuoyan Cui, Yufeng Wang, Wenxuan Shen, Chunhua Shan, Shiguang Zhang, Zhaoxiang Wang, Xinlong |
| author_facet | Deng, Haoge Pan, Ting Zhang, Fan Liu, Yang Luo, Zhuoyan Cui, Yufeng Wang, Wenxuan Shen, Chunhua Shan, Shiguang Zhang, Zhaoxiang Wang, Xinlong |
| contents | Continuous-space video generation has advanced rapidly, while discrete approaches lag behind due to error accumulation and long-context inconsistency. In this work, we revisit discrete generative modeling and present Uniform discRete diffuSion with metric pAth (URSA), a simple yet powerful framework that bridges the gap with continuous approaches for the scalable video generation. At its core, URSA formulates the video generation task as an iterative global refinement of discrete spatiotemporal tokens. It integrates two key designs: a Linearized Metric Path and a Resolution-dependent Timestep Shifting mechanism. These designs enable URSA to scale efficiently to high-resolution image synthesis and long-duration video generation, while requiring significantly fewer inference steps. Additionally, we introduce an asynchronous temporal fine-tuning strategy that unifies versatile tasks within a single model, including interpolation and image-to-video generation. Extensive experiments on challenging video and image generation benchmarks demonstrate that URSA consistently outperforms existing discrete methods and achieves performance comparable to state-of-the-art continuous diffusion methods. Code and models are available at https://github.com/baaivision/URSA |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2510_24717 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Uniform Discrete Diffusion with Metric Path for Video Generation Deng, Haoge Pan, Ting Zhang, Fan Liu, Yang Luo, Zhuoyan Cui, Yufeng Wang, Wenxuan Shen, Chunhua Shan, Shiguang Zhang, Zhaoxiang Wang, Xinlong Computer Vision and Pattern Recognition Continuous-space video generation has advanced rapidly, while discrete approaches lag behind due to error accumulation and long-context inconsistency. In this work, we revisit discrete generative modeling and present Uniform discRete diffuSion with metric pAth (URSA), a simple yet powerful framework that bridges the gap with continuous approaches for the scalable video generation. At its core, URSA formulates the video generation task as an iterative global refinement of discrete spatiotemporal tokens. It integrates two key designs: a Linearized Metric Path and a Resolution-dependent Timestep Shifting mechanism. These designs enable URSA to scale efficiently to high-resolution image synthesis and long-duration video generation, while requiring significantly fewer inference steps. Additionally, we introduce an asynchronous temporal fine-tuning strategy that unifies versatile tasks within a single model, including interpolation and image-to-video generation. Extensive experiments on challenging video and image generation benchmarks demonstrate that URSA consistently outperforms existing discrete methods and achieves performance comparable to state-of-the-art continuous diffusion methods. Code and models are available at https://github.com/baaivision/URSA |
| title | Uniform Discrete Diffusion with Metric Path for Video Generation |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2510.24717 |