search

PRBonn / lidar-bonnetal

Semantic and Instance Segmentation of LiDAR point clouds for autonomous driving
http://semantic-kitti.org
MIT License
961 stars 206 forks source link

Why permute proj xyz when process data #12

Closed hagianga21 closed 5 years ago

hagianga21 commented 5 years ago

Hi, in file parser.py, why did you permute x,y,z into z, x, y (proj_xyz.clone().permute(2, 0, 1)) ?

proj = torch.cat([proj_range.unsqueeze(0).clone(), proj_xyz.clone().permute(2, 0, 1), proj_remission.unsqueeze(0).clone()])

hagianga21 commented 5 years ago

Sorry, my mistake. I fixed it.