IOSDA

An unsupervised domain adaptation with input and output space alignment for joint optic disc and cup segmentation

---

Architecture of IOSUDA

Train

train a new model:

python train.py

---

Test

test the trained model:

python test.py

---

Metric

Left subfigure is jaccard index of OC and right subfigure is jaccard index of OD on RIM-ONE_r3 dataset.

---

BibTeX Citation

@article{chen2021iosuda,
  title={IOSUDA: an unsupervised domain adaptation with input and output space alignment for joint optic disc and cup segmentation},
  author={Chen, Chonglin and Wang, Gang},
  journal={Applied Intelligence},
  volume={51},
  number={6},
  pages={3880--3898},
  year={2021},
  publisher={Springer}
}