Active-TANGO

Quick annotation of cells - correct ML annotations

User Guide
Batches:

A batch is a collection of Machine Learning (ML) and Active annotations for a set of images.


To view a batch:

First navigate to the batches page to view all batches that you have access to. On this page, you can see some details for these batches, and can also edit and delete them. By default, the batches are ordered by the date that they were created. To change this, click on one of the other columns to order by it.

To view the ML and active annotations for a specific batch, click on the batch name. This will take you to the page for that batch. There are tabs at the top of the page to switch between ML and Active annotations. To view any of the annotations, just click on the patch name. This will take you to the dotting interface for that annotation.


To upload a new batch:

Click on the button on the home screen or in the navigation bar to go to the page. Next, fill out the name and description. Then, select the allowed users from the list, and select the zip file on your computer containing the ML annotations and image files.

Batch Zip File structure:
batch_name.zip
  • csv (ML annotation csv files)
  • images (patch image files)
  • config.csv (batch config)

Note: Annotation csv files must have the exact same name as their corresponding patch image files.

Config File example:
type,label,colour
1,Cell 1,red
2,Cell 2,green
3,Cell 3,yellow
4,Cell 4,blue
5,Cell 5,purple
6,Cell 6,cyan
7,Cell 7,pink
8,Cell 8,orange
9,Cell 9,grey

Colour options: red, green, yellow, blue, purple, cyan, pink, orange, grey

Note: There is a maximum of 9 different cell types.

Finally, click on the button.


To download annotations for a batch:

Switch to the tab and click on the button. This will download a ZIP file to your computer containing all of the active annotation JSON files.


To view the summary for a batch:

First navigate to the batch page. Then click on the button. This will bring up tables and graphs with relevant data for that batch.