Replies: 1 comment
-
Hello, It is expected behavior. Namely when sampling random location of the patch, the -1 is given to all places in segmentation map that fall outside the patch. Check script dataloader_3d.py in nnuntv2/training/dataloading. I hope this helped. |
Beta Was this translation helpful? Give feedback.
0 replies
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
-
I inspected my dataset after running nnUNets preprocessing steps. At some segmentation files there is added label with value -1. My dataset only has labels 0 for background and label 1 for the lesion of interest. Is this -1 label expected behaviour or is this some kind of error during the preprocessing?
Beta Was this translation helpful? Give feedback.
All reactions