Posts

Showing posts with the label sparse-observations

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