: The model can extend a single shot to minute-long durations by auto-regressively generating 10-second segments that seamlessly connect without visible cuts.
The video filename lu-VEnnv2a4.mp4 corresponds to a demonstration for , a training paradigm designed to improve the generation of long-form, multi-shot AI videos. lu-VEnnv2a4.mp4
According to the research published on arXiv , the key "long features" enabled by this model include: : The model can extend a single shot
: By using context-causal attention, the model supports efficient long-form generation, making it practical for creating extended visual narratives. the model supports efficient long-form generation
[2503.10589] Long Context Tuning for Video Generation - arXiv
: LCT expands the context window of video diffusion models, allowing them to maintain visual and dynamic consistency across multiple different shots within a single scene.