You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Hi, great work!
But i am confused that why do u encode motion using motion VQVAE ? As far as i know ,many works aims to generate motion and it is convenient to use the decoder of VQVAE . But MotionLLM is an UNDERSTANDING model , so is that any reason to implement the VQVAE when encoding the motion modality?
The text was updated successfully, but these errors were encountered:
Hi, great work!
But i am confused that why do u encode motion using motion VQVAE ? As far as i know ,many works aims to generate motion and it is convenient to use the decoder of VQVAE . But MotionLLM is an UNDERSTANDING model , so is that any reason to implement the VQVAE when encoding the motion modality?
The text was updated successfully, but these errors were encountered: