A collection of UNet and hybrid architectures in PyTorch for 2D and 3D Biomedical Image segmentation
A collection of UNet and hybrid architectures for 2D and 3D Biomedical Image segmentation, implemented in PyTorch.
This repository contains a collection of architectures used for Biomedical Image Segmentation, implemented on the BraTS Brain Tumor Segmentation Challenge Dataset. The following architectures are implemented
First, apply for access the BraTS Tumor dataset, and place the scans in a Data/
folder, divided into Train
and Test
. Feel free to modify the BraTS PyTorch dataloaders in data.py
for your use.
main.py
, type --help
for information on arguments.
Example: python main.py --train --cuda --data-folder "./Data/"
main_small.py
, and use --help
main_bdclstm.py
and use the weights for either your trained UNet or Small-UNet models (--help
is your savior).