I know this project is long way off (given the todo list), however would like to comment
Could this not be a dedicated scan section/button for the Digikam tag tab?
As of now one of the things seen missing is automatic scene/object detection (which you have robustly solved with this repo).
Could this idea semi-smoothly be implemented into digikam natively using the YOLOv5 (where face detection already uses V3) for object tagging with only minor adjustments to their UI? Therefore, not needing a separate python cmd/terminal script running.
I know this project is long way off (given the todo list), however would like to comment
Could this not be a dedicated scan section/button for the Digikam tag tab? As of now one of the things seen missing is automatic scene/object detection (which you have robustly solved with this repo).
Could this idea semi-smoothly be implemented into digikam natively using the YOLOv5 (where face detection already uses V3) for object tagging with only minor adjustments to their UI? Therefore, not needing a separate python cmd/terminal script running.