Skip to content

Latest commit

 

History

History
20 lines (13 loc) · 644 Bytes

README.md

File metadata and controls

20 lines (13 loc) · 644 Bytes

Python script

This folder contains Jupyter notebooks to help you train your own models and analyse your measurements.

Dependencies version

The scripts have the following dependencies and were tested with the following versions:

  • tensorflow: 1.13.1
  • keras: 2.3.1
  • seaborn: 0.10.0
  • eli5: 0.10.1
  • pandas: 1.0.3

Segmentation example

StarDist provide examples to train their model here.

Our model was trained on 100 epochs, with 48 rays, images of 320x320 pixels and with a dropout of 0.5. With GPU enabled it should take about 30 minutes to train on Google Colab.