Implement Vision related CoreML samples. So far, implemented all the models that came up new in 2019. (There is a BERT one, but it is not a Vision sample, so I skipped here.)
Detects everything pretty accurately.
Speed is better, but you can see, hand is not detected it