I've majorly followed quality.py implementation to calculate a grasp matrix to compute other grasp quality metrics (e.g., Ferrari Canny L1 metric). However, after searching other papers and documents, I wonder how the implementation of the grasp matrix function is done in this code base.
For example, when we assume 2 soft finger contacts on an object, the dimension of a grasp matrix is supposed to be 6 x 8 while the implemented function outputs 6x6 matrix, instead.
I appreciate if someone can explain a bit more about this specific implementation.
Hi,
I've majorly followed quality.py implementation to calculate a grasp matrix to compute other grasp quality metrics (e.g., Ferrari Canny L1 metric). However, after searching other papers and documents, I wonder how the implementation of the grasp matrix function is done in this code base.
For example, when we assume 2 soft finger contacts on an object, the dimension of a grasp matrix is supposed to be 6 x 8 while the implemented function outputs 6x6 matrix, instead.
I appreciate if someone can explain a bit more about this specific implementation.