This research proposes a PDE-based trainable refinement module for tubular-structure segmentation to enhance shape continuity.

- PDE-based refinement improves mask continuity while maintaining thin shape
- Our approach is applied to various architectures for tubular-structure segmentation because of post-process
- Ubuntu 24.04.2 LTS
- NVIDIA TITAN RTX(Memory: 24GB) * 4
- setup the environment.
$ pip install -r requirements.txt-
Download DRIVE dataset to
/dataset. -
Run
/dataset/data_process.pyto fit dataset for training models.
$ cd dataset
$ python data_process.py -dp DATASET_PATH -dn DATASET_NAME- train and test model.
$ cd pde-shape-refiner
$ main_ddp.sh