Rotated detection is no longer supported as a dedicated workflow in Geti v3.0

[yvolokitin]

### We want to inform the community about a change to supported workflows in the current release.

Geti™ v3.0 does not support rotated detection as a task. If you have a "rotated detection" project in Geti™ v2, we recommend migrating it to an "instance segmentation" project in v3.

Migrating a rotated detection project to v3

1. Export the dataset from your Geti™ v2 instance in COCO format.
2. In Geti v3, choose "Create new project from dataset".
3. Upload the zip archive.
4. Select "Instance segmentation" as task type.
5. Click Create and wait for the import to finish.
6. Verify that all media and their annotations are correctly imported.
7. Reinterpreting predictions as rotated bounding boxes

Since the new project is of type "instance segmentation", any trained model will produce polygons as output. At deployment time, if you need to reinterpret the predicted polygons as rotated bounding boxes, you can take advantage of the minimum_rotated_rectangle function offered by the Shapely library.

📖 For the full step-by-step guide, see the v2 → v3 migration documentation

Feedback
If rotated detection is important to your workflow, please share your use case in the comments so we can factor it into future planning.