Hello, and thank you for sharing your excellent work on MimicMotion!
I’m currently studying the hand enhancement component described in your paper and would like to understand more about the loss calculation implementation (specifically, how the hand enhancement loss is computed and integrated during training).
Would it be possible for you to release or provide a short demo snippet showing how this loss is calculated in your codebase? It would be very helpful for reproducing and understanding this part of your method.
Thank you.
Best regards,
Sang