Home

Awesome

nuclei_segmentation

Requirements

Training and Testing

First run:

python prepare_dataset.py 

to prepare dataset

python run_training.py

At the begining of the training, a folder name "test1" will be created in "experiment". During the training, models will be saved in it.

python run_training.py

It will generate a folder name '1' in 'test1'. The predictions will be shown in it. You may modify configuration.txt to change the experiement settings.

Licence

The code is licensed under MIT. Copyright (c) 2018. The dataset is obtained from https://nucleisegmentationbenchmark.weebly.com/dataset.html.