July 11, 2019 | Using AI to hide objects
Season 4 Episode 23 | 26m 46s
Engineers at Southwest Research Institute are using a technique called adversarial learning that pits artificial intelligence against itself to produce something called patches, which mask what an object really is. Engineers ask the AI program what, say, a truck would look like if it was really a bicycle. Results have military and security implications.
Aired: 07/11/19Video has closed captioning.
Problems Playing Video?
Support provided by viewers like you and the 80/20 Foundation.