Mutual-Prototype Adaptation for Cross-Domain Polyp Segmentation (JBHI 2021)

This is the official PyTorch implementation of MPA (Mutual-Prototype Adaptation) (JBHI 2021).

Mutual-Prototype Adaptation for Cross-Domain Polyp Segmentation[Paper]

Chen Yang, Xiaoqing Guo, Meilu Zhu, Bulat Ibragimov, Yixuan Yuan

Install dependencies

pip install -r requirements.txt

Datasets Preparation

CVC-DB and ETIS-Larib

(1) Download the CVC-DB and ETIS-Larib dataset.

(2) Put the data in the corresponding folders. The dataset files are organized as follows.

MPA-DA
├── data
│   ├── CVC-DB
│   │   ├── images
│   │   │   ├── [case_id].png
│   │   ├── labels
│   │   │   ├── [case_id].png
│   ├── ETIS-Larib
│   │   ├── images
│   │   │   ├── [case_id].png
│   │   ├── labels
│   │   │   ├── [case_id].png

(3) Split dataset into training set and test set as follows.

python preprocess.py

Quickstart

python ./tools/train.py

Cite

If you find our work useful in your research or publication, please cite our work:

@article{yang2021mutual,
  title={Mutual-Prototype Adaptation for Cross-Domain Polyp Segmentation},
  author={Yang, Chen and Guo, Xiaoqing and Zhu, Meilu and Ibragimov, Bulat and Yuan, Yixuan},
  journal={IEEE Journal of Biomedical and Health Informatics},
  year={2021},
  publisher={IEEE}
}