I am working on a project intended to recognize vehicles as they drive by a street and record them as they come into view. In order to do this, I am using both MobileNetDetectionNetwork and ObjectTracker nodes, with the input of a color camera (OAK-1 Lite) feeding into the detection network, which goes into the object tracker.
Currently, only two models work for me: "vehicle-detection-adas-0002" and "face-detection-retail-0044". No others seem to work. I am using the blobconverter module to take models from the OpenVINO zoo.
Is there a reason why only these networks recognize their respective objects?