Hi, thanks for the great work.
I’m wondering whether there is any plan to open-source the implementations of other evaluation metrics mentioned in your paper, including:
Generation: Subject Fidelity, Cross-View Consistency
Action-Following
Downstream Tasks: Map Segmentation, 3D Object Tracking
Thanks!