On device machine learning with Mediapipe

Machine learning models today can be consumed in two ways, either through an API or on device. Most of the time it is used through an API which means users needs to send information to a server to get a result. This could lead to privacy issues but also results taking time to return. On device machine learning solves by having a models that’s bundled with your app to do all the inference work but also keeping your input/data safe on device. In this talk we will show how we can use Mediapipe a machine learning library to implement your own custom models and common user cases in your android app. This will include Image, text and audio classifications as well as more advanced problems with a low code solution.