/UN-SAM

Primary LanguagePythonMIT LicenseMIT

Universal Prompt-Free Segmentation for Generalized Nucleus Images (UN-SAM)

This repository is an official PyTorch implementation of the paper "UN-SAM: Universal Prompt-Free Segmentation for Generalized Nucleus Images" [paper] submitted to IEEE Transactions on Medical Imaging.

Dependencies

  • Python 3.10
  • PyTorch >= 1.10.0
  • albumentations 1.5.2
  • monai 1.3.0
  • pytorch_lightning 1.1.0

Code

Clone this repository into any place you want.

git clone https://github.com/CUHK-AIM-Group/UNSAM.git
cd UNSAM
mkdir data; mkdir pretrain;

Quickstart

  • Train the UN-SAM with the default settings:
python train.py --dataset data/$YOUR DATASET NAME$ --sam_pretrain pretrain/$SAM CHECKPOINT$

Cite

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

@article{chen2024sam,
  title={UN-SAM: Universal Prompt-Free Segmentation for Generalized Nuclei Images},
  author={Chen, Zhen and Xu, Qing and Liu, Xinyu and Yuan, Yixuan},
  journal={arXiv preprint arXiv:2402.16663},
  year={2024}
}

Acknowledgements