VideoSSM: Autoregressive Long Video Generation with Hybrid State-Space Memory
This addresses the challenge of generating coherent, interactive long videos for applications like streaming and content creation, representing a novel method for a known bottleneck.
The paper tackled the problem of maintaining coherence in autoregressive long-video generation by proposing VideoSSM, which unifies AR diffusion with a hybrid state-space memory, achieving state-of-the-art temporal consistency and motion stability at minute-scale horizons.
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.