DistNet

page under construction

Deep Bacteria Segmentation and Tracking with DistNet: Bacmman Module

This tutorial is based on the work described in this publication

Download example dataset and configure

1. If necessary, install bacmman.
2. Download images from example dataset 1
3. If using bacmman for the first time: choose a working directory through right-click on the panel below Working Directory
4. From bacmman software, choose menu command: Dataset>New Dataset from github , set jeanollion as username in the Github Credentials panel, and select the configuration ExampleDatasets>dataset1_distnet
5. Download the trained model: choose menu command Export>DL Model library, set jeanollion as username in the Github Credentials panel. In the DL Models panel double-click on the model DistNet>DistNet_base>URL to download it. Unzip it and place it in a folder named model_bact_phase in the working directory (note: this directory can be changed in the configuration of the DLEngine the object class's tracker).

Run processing

The processing pipeline is the following:

1. Images are cropped during pre-processing
2. Microchannels are segmented
3. Bacteria are segmented and tracked in each microchannel

To run step 1) and 2): from the Home tab select Pre-Processing & Segment and Track in the Task panel and Microchannels in the Objects panel.

See this page to adapt DistNet to a dataset