Object detection in videos

apart from not being maintained from opencv, – why so ?
if there are problems, describe in more detail, please.

what exactly ? are those covered by the 80 COCO classes ?