VideoSSM: Autoregressive Long Video Generation with Hybrid State-Space Memory
Fuente:
arXiv
Guardado en:
| Autores principales: | , , , , , , , , , , |
|---|---|
| Formato: | Preprint |
| Publicado: |
2025
|
| Materias: | |
| Acceso en línea: | |
| Etiquetas: |
Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
|
| _version_ | 1866915653357666304 |
|---|---|
| author | Yu, Yifei Wu, Xiaoshan Hu, Xinting Hu, Tao Sun, Yangtian Lyu, Xiaoyang Wang, Bo Ma, Lin Ma, Yuewen Wang, Zhongrui Qi, Xiaojuan |
| author_facet | Yu, Yifei Wu, Xiaoshan Hu, Xinting Hu, Tao Sun, Yangtian Lyu, Xiaoyang Wang, Bo Ma, Lin Ma, Yuewen Wang, Zhongrui Qi, Xiaojuan |
| contents | Autoregressive (AR) diffusion enables streaming, interactive long-video generation by producing frames causally, yet maintaining coherence over minute-scale horizons remains challenging due to accumulated errors, motion drift, and content repetition. We approach this problem from a memory perspective, treating video synthesis as a recurrent dynamical process that requires coordinated short- and long-term context. We propose VideoSSM, a Long Video Model that unifies AR diffusion with a hybrid state-space memory. The state-space model (SSM) serves as an evolving global memory of scene dynamics across the entire sequence, while a context window provides local memory for motion cues and fine details. This hybrid design preserves global consistency without frozen, repetitive patterns, supports prompt-adaptive interaction, and scales in linear time with sequence length. Experiments on short- and long-range benchmarks demonstrate state-of-the-art temporal consistency and motion stability among autoregressive video generator especially at minute-scale horizons, enabling content diversity and interactive prompt-based control, thereby establishing a scalable, memory-aware framework for long video generation. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2512_04519 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | VideoSSM: Autoregressive Long Video Generation with Hybrid State-Space Memory Yu, Yifei Wu, Xiaoshan Hu, Xinting Hu, Tao Sun, Yangtian Lyu, Xiaoyang Wang, Bo Ma, Lin Ma, Yuewen Wang, Zhongrui Qi, Xiaojuan Computer Vision and Pattern Recognition Autoregressive (AR) diffusion enables streaming, interactive long-video generation by producing frames causally, yet maintaining coherence over minute-scale horizons remains challenging due to accumulated errors, motion drift, and content repetition. We approach this problem from a memory perspective, treating video synthesis as a recurrent dynamical process that requires coordinated short- and long-term context. We propose VideoSSM, a Long Video Model that unifies AR diffusion with a hybrid state-space memory. The state-space model (SSM) serves as an evolving global memory of scene dynamics across the entire sequence, while a context window provides local memory for motion cues and fine details. This hybrid design preserves global consistency without frozen, repetitive patterns, supports prompt-adaptive interaction, and scales in linear time with sequence length. Experiments on short- and long-range benchmarks demonstrate state-of-the-art temporal consistency and motion stability among autoregressive video generator especially at minute-scale horizons, enabling content diversity and interactive prompt-based control, thereby establishing a scalable, memory-aware framework for long video generation. |
| title | VideoSSM: Autoregressive Long Video Generation with Hybrid State-Space Memory |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2512.04519 |