Microscopy U-Net

Description

Source code for the kaggle Data Science Bowl 18 competition. Includes a Mask-RCNN (from Matterport) and a U-Net model.

Run the testing by starting test_mask_rcnn.py (Python 3.6, Tensorflow 1.6). Predicted masks will be saved in data/stage*_test/[img]/masks_predicted.

Name		Name	Last commit message	Last commit date
Latest commit History 186 Commits
models		models
.gitignore		.gitignore
LICENSE		LICENSE
README.md		README.md
elastic_transform.py		elastic_transform.py
encode_submissions_mask_rcnn.py		encode_submissions_mask_rcnn.py
encode_submissions_unet.py		encode_submissions_unet.py
ensemble_mask_rcnn.py		ensemble_mask_rcnn.py
eval_mask_rcnn.py		eval_mask_rcnn.py
hyperparam_optim_mask_rcnn.py		hyperparam_optim_mask_rcnn.py
kmeans.py		kmeans.py
preprocess_masks.py		preprocess_masks.py
remove_predicted.py		remove_predicted.py
requirements.txt		requirements.txt
startup.sh		startup.sh
test_mask_rcnn.py		test_mask_rcnn.py
test_separate.py		test_separate.py
test_unet.py		test_unet.py
train_mask_rcnn.py		train_mask_rcnn.py
train_separate.py		train_separate.py
train_unet.py		train_unet.py