This repo contains the Pytorch implementation of our paper:
Yu Tian, Guansong Pang, Fengbei Liu, Seon Ho Shin, Johan W Verjans, Rajvinder Singh, Gustavo Carneiro.
- Accepted at MICCAI 2021.
Please download the Hyper-Kvasir Anomaly Detection Dataset from this link.
The code is build based on the SCAN.
Modify the dataloader (data/lag_loader.py) code for your own medical images, then simply run the following command:
python simclr.py --config_env configs/env.yml --config_exp configs/pretext/simclr_cifar10.yml
If you find this repo useful for your research, please consider citing our paper:
@inproceedings{tian2021constrained,
title={Constrained contrastive distribution learning for unsupervised anomaly detection and localisation in medical images},
author={Tian, Yu and Pang, Guansong and Liu, Fengbei and Chen, Yuanhong and Shin, Seon Ho and Verjans, Johan W and Singh, Rajvinder and Carneiro, Gustavo},
booktitle={International Conference on Medical Image Computing and Computer-Assisted Intervention},
pages={128--140},
year={2021},
organization={Springer}
}