A concept for paintings that evolve through the process of interactive selection using gaze tracking.
At Nabi AI Hackathon 2016 in Seoul I presented a working demo for genetic algorithmic paintings that evolve through the process of interactive selection using gaze tracking. Or to put it more simple: A set of paintings that evolve based on which ones people stare the most at.
How does it work?
The demo is built in p5.js with an eye tracker from The Eye Tribe controlling the fitness of the paintings via a websocket connection. In the online demo version this interaction is replaced by the position of the mouse.
Code for this project is available at https://github.com/AndreasRef/geneticPainting.