‘Seeing Things’ is a collaborative project between code developer Harry Van Ness and sound designer Daryosh Karimi. It aims to use a class of machine learning, called generative adversarial networks (GAN), an idea originally designed by Ian Goodfellow. It involves two neural networks, the ‘Creator’ and the ‘Discriminator’. They duel one another and by fooling each other on a random basis, they visualise new imagery, redefining the word ‘real’.
Original code developed at Google by Brock et al. (2018)
You're on the go, it's understandable. (This website works best only on Desktop.)