InstanceSegmentation-CycleGAN

An instance segmentation network for unsupervised domain adaptation based on CycleGAN.

Task overview

Model architecture

Files

You can use https://nbviewer.jupyter.org/ and paste the .ipynb link into it if Girhub does not render the file properly

To train ISC-GAN:

• Train seg-GAN-curriculum step1.ipynb to segment fake target image

• Train seg-GAN-curriculum step2.ipynb to segment real target image

To evaluate:

• Evaluate seg-GAN-curriculum.ipynb

Requirements:

matplotlib==3.0.3

numpy==1.17.2

opencv-python==3.2.0.7

pycocotools==2.0.0

scikit-image==0.15.0

scipy==1.2.1

torch==1.4.0

torchvision==0.5.0

Citation:

Wang S, Rong R, Gu Z, Fujimoto J, Zhan X, Xie Y, Xiao G. Unsupervised Domain Adaptation for Nuclei Segmentation: Adapting from Hematoxylin & Eosin Stained Slides to Immunohistochemistry Stained Slides Using a Curriculum Approach. Computer Methods and Programs in Biomedicine. 2023 Aug 19:107768.