Posts

Showing posts with the label full-body-reconstruction

Disentangled Motion Representation: Encoding Full-Body Avatars into Discrete Latent Spaces