search

JonasSchult / Mask3D

Mask3D predicts accurate 3D semantic instances achieving state-of-the-art on ScanNet, ScanNet200, S3DIS and STPLS3D.
MIT License
545 stars 110 forks source link

Training on ScanNet200: data prepare rRaise an exception #78

Closed xiaotiancai899 closed 1 year ago

xiaotiancai899 commented 1 year ago

Hi Jonas! This is a fantastic job. When I train on the scannet dataset, I met a problem about the dataset prepare module. In semseg.py line 194, when the code runs to "self._labels = self._select_correct_labels(labels, num_labels)", there exists an exception. From another issue I know I created the ScanNet200 dataset (with 200 labels) but wanted to train a model for the original ScanNet dataset with 20 labels) on it. But I do not know how to verify in my preprocessing scripts if I enabled scannet200. Could you please show me the detail about it? Thanks a lot！！！

File "/mnt/d/student/Documents/software/wsl/mask3d/mask3d_main/mask3d_main/datasets/semseg.py", line 194, in init self._labels = self._select_correct_labels(labels, num_labels) File "/mnt/d/student/Documents/software/wsl/mask3d/mask3d_main/mask3d_main/datasets/semseg.py", line 595, in _select_correct_labels raise ValueError(msg) ValueError: not available number labels, select from: 200, 200

JonasSchult commented 1 year ago

Hi!

Thanks for your interest in our work! :)

You can set --scannet200=false to create a dataset with 20 labels (see here).

Best, Jonas