Anyone found out if it‘s possible to add bounding box annotations to an image training set similar to the turicreate approach?
https://github.com/apple/turicreate/tree/master/userguide/object_detection
If not, this would be a major limitation for an otherwise great tool.
