Video Depth without Video Models
Quickly edit the expression of a face
LargeSpatialModel: End-to-end Unposed Images to Semantic 3D