Ruining Li

New models enhance how objects respond to user actions through visual interfaces.

A new model generates realistic videos showing part-level movements from single images.

New methods create realistic interactions between digital humans and objects using text descriptions.