search

mlexchange / mlex_highres_segmentation

A Dash interface for ML-based segmentation of user-annotated large multi-dimensional image data
Other
5 stars 4 forks source link

Retrieve segmentation results and populate job parameters #178

Closed Wiebke closed 7 months ago

Wiebke commented 7 months ago

This PR adds the ability to retrieve segmentation results and populates job parameters based on model parameters and currently selected data.

To test this without a segmentation model, but functional jobs (e.g. using the sleep command instead of an actual train command):

  1. Annotate a selected data set and hit Train. This will save a mask. Note down the mask_uri (printed in the notification)
  2. Additionally, note down the flow uid of the first child flow of the training flow (see Details of the flow in the Prefect UI)
  3. Create a copy of the mask into a container under SEG_TILED_API

The process for testing the inference retrieval is similar, but requires a data set of the same size as the original data. See example/copy_mask_as_result.py

Wiebke commented 7 months ago

Closes #175.