Add start and end time to run object detection in the body of the request. Backend needs to take included clips into account.