Inquiry About the Design of the Hand-Object Interaction (HOI) Coordinate System i
In the current implementation, a third coordinate system, referred to as the Hand-Object Interaction (HOI) coordinate system is introduced to predict the SDF of the object before transforming it to the hand coordinate system.
I have one question regarding the design choice of introducing this additional interaction coordinate system:
Why is the interaction coordinate system necessary? What is the motivation for introducing this intermediate coordinate system instead of directly predicting the object's SDF in the hand coordinate system?
One reason I think is predicting SDF in the intermediate system makes the sampling 3D point in [-1, 1] possible, which cannot be done if you predict in the hand system.
Looking for your insights.