I am training / using OpenCV to try and discover human figures. I want to try to train HOG for my specific goals and not use the provided getDefaultPeopleDetector function. I could not find any useful documentation in the HOGDescriptor class.
How do I prepare my classifier for my own purposes?
source share