A Berlin artist heading a project to trick facial recognition through fashion, has said the increased use of such tools to profile people on physical traits reminds him of eugenics.
Adam Harvey, in collaboration with New York studio Hyphen Labs, has put forward a way to throw off facial recognition with specially designed clothing and textiles.
The project, dubbed Hyperface, has been described as a “new kind of camouflage” to reduce the confidence score of facial detection.
Basically, Hyperface operates in an ‘I’m Spartacus’ way, in that instead of concealing your face, it overloads facial recognition programmes with multiple false faces through patterns printed on clothing.
Introducing the concept, Harvey said: “As I’ve looked at in an earlier project, you can change the way that you appear.”
He explained that Hyperface involved the creation of textile patterns that “oversaturated an area with faces to divert the gaze of a computer recognition algorithm.
“It could be used to hopefully modify the environment around you, whether it is somebody next to you, whether you’re wearing it.”
Harvey recently spoke about Hyperface at the Chaos Communication Congress in Hamburg, Germany.
Listing “crazy” ways in which facial detection tools are being developed to profile people, he said: “What all this reminds me of is of Francis Galton and eugenics. Who the real criminal in these cases would be the people who are perpetrating this idea and not the people who are being looked at.”
During the talk based on “hyper-competitive retail surveillance”, the designer outlined how Hyperface aimed to add to the discourse on facial recognition technology while also altering its success rate.
Through another project known as CV Dazzle, named after a WWI camouflage tactic for warships, Harvey previously used hairstyles and patterned facial designs to exploit vulnerabilities in facial recognition technology.
“By doing hair and makeup in a certain location you decrease the confidence score that [your] face would appear [on a recognition program],” he explained to the audience at the Chaos Communication Congress.