Hey, So I was working on a project where I wanted be able to identify certain images using features. I was planning to use the OpenCV feature detection and then be able to port that whole algorithm to a Keras model so that I can run it on CoreML for iOS with pretty good frame rates. Thanks!