Model co-generates video and 3D motion for arbitrary rigs
Cross-attention links a skeleton motion model to a pretrained video model, trained on nearly 200K pairs.
Stories per day · 30 days
1 stories, newest first · one card per story
Stories tagged #character-rigs