SoftCut: a fully differentiable relaxed graph cut approach for deep learning image segmentation

Repository for the experiments scripts for the LOD 2023 of the paper SoftCut: a fully differentiable relaxed graph cut approach for deep learning image segmentation

Python requirements

- pytorch>=1.8
- cudatoolkit
- neptune-client
- pytorch-lightning
- torchvision>=0.9
- torchaudio
- psutil
- scikit-learn
- scikit-image
 - pybind11>=2.2
- omegaconf
- segmentation-models-pytorch=0.2.0
- prox-tv
- torch-scatter
- albumentations
- suitesparse
- scipy
- torch_sparse_solve
- torch-submod (https://github.com/marcocannici/torch-submod.git)