Skip to content

[TMI' 21] DSI-Net: Deep Synergistic Interaction Network for Joint Classification and Segmentation with Endoscope Images

Notifications You must be signed in to change notification settings

CityU-AIM-Group/DSI-Net

Repository files navigation

DSI-Net

This repository is an official PyTorch implementation of the paper "DSI-Net: Deep Synergistic Interaction Network for Joint Classification and Segmentation with Endoscope Images", TMI 2021.

Dependencies

  • Python 3.6
  • PyTorch >= 1.3.0
  • numpy
  • apex
  • sklearn
  • matplotlib
  • PIL

Usage

  • Downloading processed dataset
  • Training DSI-Net
    python train_DSI_Net.py --gpus 0 --K 100 --alpha 0.05 --image_list 'data/WCE/WCE_Dataset_image_list.pkl'

Citation

@ARTICLE{9440441,
  author={Zhu, Meilu and Chen, Zhen and Yuan, Yixuan},
  journal={IEEE Transactions on Medical Imaging}, 
  title={DSI-Net: Deep Synergistic Interaction Network for Joint Classification and Segmentation with Endoscope Images}, 
  year={2021},
  doi={10.1109/TMI.2021.3083586}}

Contact

Meilu Zhu ([email protected])

About

[TMI' 21] DSI-Net: Deep Synergistic Interaction Network for Joint Classification and Segmentation with Endoscope Images

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published