Tom White: Neural Abstractions
How do machines perceive the world? Tom White has investigated enabling computer vision systems to draw their own visual abstractions through the creation of abstract ink print artworks. Though these artificial vision systems are trained only on real world images, when forced to express themselves abstractly they are able to create simpler forms that match their internal representations. These prints are also understood universally across most AI systems trained to recognize the same objects. It’s art by AI, for AI. By giving the algorithms the ability to express themselves directly, we are better able to see the world through the eyes of a machine.
Tom White is a senior lecturer in computational design at Victoria University of Wellington School of Design, where he teaches creative coding and researches creative applications of modern AI techniques.
For more information, visit artsdesign.berkeley.edu.
Presented by the Berkeley Center for New Media; cosponsored by Autolab and CITRIS People and Robots (CPAR).