dingluxin/udc-net

Code for Dual-Consistency Semi-Supervised Learningwith Uncertainty Quantification for COVID-19Lesion Segmentation from CT Images

UDC-Net

This is a implementation of the paper "Dual-Consistency Semi-Supervised Learning with Uncertainty Quantification for COVID-19 Lesion Segmentation from CT Images ".

pdf link of the paper: https://arxiv.org/abs/2104.03225

@article{li2021dual,
  title={Dual-Consistency Semi-Supervised Learning with Uncertainty Quantification for COVID-19 Lesion Segmentation from CT Images},
  author={Li, Yanwen and Luo, Luyang and Lin, Huangjing and Chen, Hao and Heng, Pheng-Ann},
  journal={arXiv preprint arXiv:2104.03225},
  year={2021}
}

Installation

This repository is based on Pytorch 1.3.0

Usage

1. Clone the repository:

git clone https://github.com/poiuohke/UDC-Net
cd UDC-Net

2. change the data path and hyper-parameters in ./configs/config.json

3. Train the model:

python train.py

4. Inference

python inference.py --config ./configs/config.json --model trained_model_path --data_path test_data_path --mask_path test-mask_path