MV-Forcing: Long Multi-View Video Generation via 4D-Grounded Spatio-Temporal Self-Forcing
Researchers introduce MV-Forcing, a framework for generating long, multi-view consistent videos. It combines temporal and view-wise autoregression using 4D geometric bridging and spatio-temporal distillation. This approach allows for temporally unbounded generation of geometrically consistent multi-view videos.
#4d-geometric-bridging#ai-research#multi-view-video-generation#spatio-temporal-distillation#video-diffusion
Read original ↗Sign in to upvoteSign in to save