Hello,
I have modified deepstream-app2 from deepstream_python_apps to work only for classification. Please, note that, I do not need object detection, and want to perform classification on entire frame.
I have removed primary detector, How can I access the predictions(outputs) of the classifier in the probe function? And draw it on the frame.
Also, these are some variables:
frame_meta.misc_frame_info : [0 0 0 0]
frame_meta.frame_user_meta_list : None
frame_meta.display_meta_list : None
frame_meta.obj_meta_list : None
frame_meta.bInferDone : 0
Thanks.