ma-xu/pointMLP-pytorch

farthest_point_sample

mahaotian-io opened this issue · 6 comments

mahaotian-io commented

I found in your pointmlp model that you have a function defined farthest_point_sample but you still use .furthest_point_sample function in the pointnet2_utils, and I want to ask, is there a difference between the two functions?

ma-xu commented

@mahaotian-io Hi, they are the same, but the self-defined function defined farthest_point_sample is in python and .furthest_point_sample is in cuda. We use the cuda version for fast training and inference.

mahaotian-io commented

I would like to ask you about the "GEOMETRIC AFFINE MODULE" in the pointmlp model specific implementation.

ma-xu commented

I would like to ask you about the "GEOMETRIC AFFINE MODULE" in the pointmlp model specific implementation.

Sure, what is ur question?

mahaotian-io commented

"GEOMETRIC AFFINE MODULE"

How do you implement "GEOMETRIC AFFINE MODULE" in your code? I want to try to use this method in some other tasks.

ma-xu commented

@mahaotian-io You can find it in this module:

pointMLP-pytorch/classification_ModelNet40/models/pointmlp.py

Line 130 in 2056698

class LocalGrouper(nn.Module):

mahaotian-io commented

@mahaotian-io You can find it in this module:

pointMLP-pytorch/classification_ModelNet40/models/pointmlp.py

Line 130 in 2056698

class LocalGrouper(nn.Module):

thank you