PointNeXt
PointNeXt copied to clipboard
Semantic3D lower result
Thanks for the great work and open-sourcing it. @guochengqian Has anyone tested PointNext on Semantic3d (reduced-8) ? It seems the results are very weak (close to mIoU 50) and model are underfitting.
Have you played with radius and data augmentations. Radius matter a lot for different datasets. Please let me know
@guochengqian thank's for the reply Have you any ideas of the possible range of radius ? With Semantic3D, I also have one other problem referenced in #138 ; concerning ignore_index in IoU introducing a gap between train and val IoU. I will try to change radius. Thank's.