LynnReal-Omni brings text-to-video, image-to-video, human- and hand-pose guided generation, structural control, omni-reference generation, style transfer, video editing, degraded-video restoration and streaming long-video generation into a single framework, all at four-step fast generation.
The paper list of the survey "The Past Frames the Future: Memory Mechanisms for Autoregressive Video Generation"