Fabian Isensee

icon location interested in deep learning with focus on medical image segmentation

Results 453 comments of Fabian Isensee

RuntimeError: One or more background workers are no longer alive. Exiting. Please check the print statements above for the actual error message

Any update?

Why do I get an error when processing? The directory structure of labelsTr should only have labels, why is there also img?

Does the expected label file exist?

Why do I get an error when processing? The directory structure of labelsTr should only have labels, why is there also img?

Any update on this?

GPU and CPU from util 100% to util 0% after 5-6 epochs.

Hey, sorry for the late reply. Issues like these are super difficult for me to debug because it only appears on some systems. If there was a code snippet that...

GPU and CPU from util 100% to util 0% after 5-6 epochs.

Hey, what's the status here?

A question on exchanging label 1 and 2 in Dataset137_Bra21.py

Nothing is overwritten and we need no temporary changes because we write to a new target array which is initialized with 0 and we read from the original segmentation. The...

Using custom architectures

Hey, so all you need to test different architectures is to overwrite build_network_architecture plus any additional changes you'd like to make. Basically what you did is exactly correct. The 'quick...

Using custom architectures

Hey, thanks for sharing! Three things to be aware of: - do not hard code the patch size in build_network_architecture! Use the given ConfigurationManager to read the patch size nnU-Net...

Using custom architectures

Hah you are right the signature is ``` def build_network_architecture(architecture_class_name: str, arch_init_kwargs: dict, arch_init_kwargs_req_import: Union[List[str], Tuple[str, ...]], num_input_channels: int, num_output_channels: int, enable_deep_supervision: bool = True) -> nn.Module: ``` I just...

Using custom architectures

Hey, how are you progressing? Is everything working now?