Adding a folder below the medigan repo "models" where downloaded and static models will be stored.
Adding the first static model to medigan: 00007_BEZIERCURVE_TUMOUR_MASK.
Static models are lightweight models (i.e. data generators without deep learning models as backbone) that can hence be stored inside medigan instead of Zenodo.
00007_BEZIERCURVE_TUMOUR_MASK generates tumour segmentation masks based on geometric operations.
This PR has two main features:
Tested with:
python -m tests.model_integration_test