ValueError: MirrorTransform now takes the axes as the spatial dimensions.

[bahetibhakti]

I installed all the required dependancies and tried hard to run the code but stuck at following error. Please help

Reading from /home/gpu3/prob_unet/output/quarter/train/stuttgart Reading from /home/gpu3/prob_unet/output/quarter/train/weimar Reading from /home/gpu3/prob_unet/output/quarter/train/jena Reading from /home/gpu3/prob_unet/output/quarter/train/aachen Reading from /home/gpu3/prob_unet/output/quarter/train/bochum Reading from /home/gpu3/prob_unet/output/quarter/train/hanover Reading from /home/gpu3/prob_unet/output/quarter/train/tubingen Reading from /home/gpu3/prob_unet/output/quarter/train/erfurt Reading from /home/gpu3/prob_unet/output/quarter/train/bremen Reading from /home/gpu3/prob_unet/output/quarter/train/zurich Reading from /home/gpu3/prob_unet/output/quarter/train/hamburg Reading from /home/gpu3/prob_unet/output/quarter/train/dusseldorf Reading from /home/gpu3/prob_unet/output/quarter/train/cologne Reading from /home/gpu3/prob_unet/output/quarter/train/strasbourg Reading from /home/gpu3/prob_unet/output/quarter/train/krefeld train set comprises 2701 files. Traceback (most recent call last): File "train_prob_unet.py", line 188, in train(cf) File "train_prob_unet.py", line 39, in train data_provider = get_train_generators(cf) File "/home/gpu3/probabilistic_unet/training/data/cityscapes/data_loader.py", line 335, in get_train_generators n_batches=cf.n_train_batches) File "/home/gpu3/probabilistic_unet/training/data/cityscapes/data_loader.py", line 300, in create_data_gen_pipeline mirror_transform = MirrorTransform(axes=(3,)) File "/home/gpu3/probabilistic_unet/training/batchgenerators/transforms/spatial_transforms.py", line 192, in init raise ValueError("MirrorTransform now takes the axes as the spatial dimensions. What previously was " ValueError: MirrorTransform now takes the axes as the spatial dimensions. What previously was axes=(2, 3, 4) to mirror along all spatial dimensions of a 5d tensor (b, c, x, y, z) is now axes=(0, 1, 2). Please adapt your scripts accordingly.

[BrianDo2005]

Hi @bahetibhakti Have you found out the solution for the above error? I encounter the same error.

[hajtos]

I know that this is a bit late for an answer, but replacing line 300 from file probabilistic_unet/training/data/cityscapes/data_loader.py from: mirror_transform = MirrorTransform(axes=(3,)) to: mirror_transform = MirrorTransform(axes=(1,)) fixes the issue