Generate a novel-view video from a single image
Retrieve 3D human motion videos from text descriptions
Generate animated videos from images and motion sequences
Generate human motion from text or audio