ShapePU
This project is developed for our MICCAI 2022 paper: ShapePU: A New PU Learning Framework Regularized by Global Consistency for Scribble Supervised Cardiac Segmentation.
Datasets
- The MSCMR dataset with mask annotations can be downloaded from MSCMRseg.
- Our scribble annotations of MSCMRseg have been released in MSCMR_scribbles.
- The scribble-annotated MSCMR dataset used for training could be directly downloaded from MSCMR_dataset.
- The ACDC dataset with mask annotations can be downloaded from ACDC and the scribble annotations could be downloaded from ACDC scribbles. Please organize the dataset as the following structure:
XXX_dataset/
-- TestSet/
--images/
--labels/
-- train/
--images/
--labels/
-- val/
--images/
--labels/
Usage
- Set the “dataset” parameter in main.py, line 55, to the name of dataset, i.e., “MSCMR_dataset”.
- Set the “output_dir” in main.py, line 57, as the path to save the checkpoints.
- Download the dataset, for example, the MSCMR_dataset. Then, Set the dataset path in /data/mscmr.py, line 110, to your data path where the dataset is located in.
- Check your GPU devices and modify the “GPU_ids” parameter in main.py, line 61. Start training with:
python main.py
Requirements
This code has been tested with Python 3.8.5 torch 1.7.0 torchvision 0.8.0
If you have any problems, please feel free to contact us. Thanks for your attention.