PH-Net

Official implementation of "PH-Net: Semi-Supervised Breast Lesion Segmentation via Patch-wise Hardness", CVPR 2024.

Notes

Due to the confidentiality agreement in commercial cooperation, we only provide codes of core modules and the whole trainable models for the convenience of comparisons.

Requirements

• python >= 3.6
• pytorch >= 1.10.2

Citation

If you find our code useful, please cite our paper.

@inproceedings{jiang2024ph,
  title={PH-Net: Semi-Supervised Breast Lesion Segmentation via Patch-wise Hardness},
  author={Jiang, Siyao and Wu, Huisi and Chen, Junyang and Zhang, Qin and Qin, Jing},
  booktitle={Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition},
  pages={11418--11427},
  year={2024}
}

Acknowledgement

We thank U²PL for part of their codes.