Automatic segmentation of zebrafish bone from uCT data using deep learning.
pip install git+https://github.com/wahabk/ctfishpy
Check the jupyter notebooks in examples/
For python dependencies check setup.py
I provide a docker image to make gpu training easier
Step 1 : Build image, this will build an image named zebrafish
docker build . --tag=zebrafish
docker build . --tag=zebrafish --network=host # if you're on vpn
Step 2 : Create container in interactive mode, this will start a shell inside the container
docker run -it \
-v <data_dir>:<data_dir> \ # mount data directory as volume
-v <repo_dir>:/colloidoscope \ # mount git repo directory as volume for interactive access
--gpus all \ # allow the container to use all gpus
--network=host \ # if you're on vpn
zebrafish \
Note: If you want to launch the container on a custom hard drive use:
sudo dockerd --data-root <custom_dir>