pointnet.pytorch
pointnet.pytorch copied to clipboard
Is rotation matrix defined correctly in dataset.py?
https://github.com/fxia22/pointnet.pytorch/blob/f0c2430b0b1529e3f76fb5d6cd6ca14be763d975/pointnet/dataset.py#L182
I understand this rotation matrix for data argumentation. Thus, theta and -theta might function the same. Logically, if we are doing pc.dot(rotation matrix), I guess we need to transpose the rotation_matrix in the original code?