Changlin Qian

How can I generate the visual perception(rgb image) of the agent at any arbitrary postions and poses?

3

My aim is to export the visual perception(rgb image or depth image) of the agent for any given position and pose.

Question about ablation study on token number

1

Excellent work! The ablation on token number M shows a great boost in performance. I would like to know whether these M tokens are used to represent SMPL parameter at...