Let say, I want to predict my input data comparing my data. Firstly, I will detect the objects in input. Then I need to group each class to then comparing with the original datasets it correct or not. How can I do in OpenCV AND/OR deep learning model?