This article was taken from the January 2012 issue of Wired magazine. Be the first to read Wired's articles in print before they're posted online, and get your hands on loads of additional content by subscribing online.
Face recognition is ubiquitous and "freaks most people out", says Adam Harvey. So the New York-based designer created CV Dazzle, a camouflage that prevents computers recognising faces.
Ignore accessories That fedora isn't going to fool anyone. Face-recognition algorithms typically focus on a small triangular area starting above the eyes and down to the chin, widest at the eyes. "That's where most of the information is -- ears don't affect it too much, nor does wearing a hat," says Harvey.
Go monochrome
"A lot of face-detection algorithms convert photos into black and white," says Harvey. "When you're applying make-up, it's really about the white and the black, the contrast." He recommends off-the-shelf black-and-white make-up which doesn't crack when it dries. Apply light foundation.
Avoid enhancers
Enhancers such as eye shadow and lipstick amplify key features. "They make your eyes darker and lips more pronounced. You want to do the opposite," Harvey says. The old adage with make-up is less is more. "You don't want to look ridiculous. Camouflage must not be thought of as a mask."
Mask the T-zone
The region where eyes, nose and forehead intersect is "probably the most focused-on area in facial recognition". But there are a few effective ways to fool computer vision. Bring hair down across the eyeline. If you have dark skin, bring down light hair, and vice versa. Or use eye shadow to disguise contours.
This article was originally published by WIRED UK
