Throughout the day we spend a lot of time interacting with technology. Mostly through presses of simple buttons. But why? This project explores possibilities to use expressive body moves to control your environment by the use of machine learning.
On the laptop incoming webcam video is processedin P5js together with a machine learning library Ml5 to recognise poses. After that it applies the K-nearest Neighbour algorithm to classify different poses. Once the position to trigger the finger is seen a signal is send to the Arduino over bluetooth to move the finger.
A live demo for audience of the project. In this case the DeepFinger is used to control Netflix with a pose.