Modified UNet for Zebrafish head and operculum segmentation
Description
- Data folder contains the test images
- Models folder contains the trained models directly used for testing
- Model architecture is defined in unet_modified.py
- Main file is run.py
- All the supporting functions are written in utils.py
- Custom loss functions used in the models are written in loss_functions.py for compiling the model
- Code is compatible with the open-source tool Cytomine ULiège R&D version (https://uliege.cytomine.org)
Reference
- Deep Learning appraoches for Head and Operculum Segmentation in Zebrafish Microscopy Images, Navdeep Kumar et. al. In the 19th International Conference on Computer Analysis of Images and Patterns (CAIP-2021)