As surveillance and security concerns rise, so does the importance of technology used to identify individuals.
Facial recognition is becoming increasingly popular these days. You see it in the movies, in Facebook, in Google Photos and what not. Facebook searches for its users’ photos to add tags to them and Amazon’s physical store scans faces to recognize you as a user.
So much that both personal and commercial usage of this technology has also been exponentially increasing these past few years.
With such increases also come some hard questions about privacy and information security.
It is exactly these privacy concerns that have prompted an artist and technologist Adam Harvey to come up with something that can completely bypass a facial recognition system.
His solution? Giving the system what it wants – faces. Actually, it gives the system lots of faces, which are printed and present on the clothes he has designed for the very purpose.
The result? Facial recognition systems are overwhelmed by thousands of faces and thus don’t work as they should.
The Hyperface Project
Deemed the “Hyperface project”, clothes or different textiles are printed with specific patterns that appear to have eyes, hair, mouths and other facial features which a computer can recognize as a face.
Image of a Hyperface Pattern
Harvey explained how it all works in a technical sense:
As I’ve looked at in an earlier project, you can change the way you appear. But, in camouflage you can think of the figure and the ground relationship. There’s also an opportunity to modify the ‘ground’, the things that appear next to you, around you, and that can also modify the computer vision confidence score.
Objects Around You Can Distract A Security System
What he meant is that you can modify objects around you which can help distract a security system’s facial recognition software to not look at your face but instead focus on the modified object. Hyperface aims to do exactly this in order to divert a computer’s attention towards what it wants – via faces.
Essentially, Hyperface can modify the environment around you. You can be wearing something with a Hyperface pattern on it, and it can be someone next to you as well.
Here’s How It Works
Adam has collected and compiled 47 different data points which can be discovered from a 100 x 100 pixel image of a face. These data points are something commercial and academic researchers use to recognize faces from images.
The different traits used to recognize faces include age, gender, “calm”, “kind” and even criminal tendencies like “pedophile” or “white collar offender”.
A number of researchers are finding ways to turn that data into an insight that can be used for marketing.
Harvey adds to this:
What all this reminds me of is Francis Galton and eugenics. The real criminal, in these cases, are people who are perpetrating this idea, not the people who are being looked at.
More details about the Hyperface project will be revealed by Harvey and Hyphen-Labs (partnering on the project) later this month.
Via The Guardian
