I've been running deepstack integration for over a year now, in the "old school" method (Using AITool and sending it jpgs), with separate camera objects for substream and main feed. The substream is the motion detection and aitool/deepstack triggers the main camera (which has no motion detection).

I'd like to migrate over to the "new school" way of doing it, with only a single camera object and direct deepstack integration. I've tried to do this on a couple of my cameras, and i keep running into the same issue:

Using UI3's auto maximize feature, it will auto maximize a camera that just detects motion... motion that is not verified as an object by deepstack. Is there a way to make this behave like my previous setup, where a camera would only maximize if deepstack determined there was an interesting object in the frame? It is canceling the alert of this camera because deepstack doesn't detect anything, but that doesn't prevent it from maximizing the camera when it detects motion.