Who Are You Looking At? Glasses Fool Face Recognition Software

Who Are You Looking At? Glasses Fool Face Recognition Software
Examples of successful impersonation and dodging attacks. Fig. (a) Shows SA (top) and SB (bottom) dodging against DNNB. Fig. (b) – (d) show impersonations. Impersonators carrying out the attack are shown in the top row and corresponding impersonation targets in the bottom row. Fig. (b) Shows SA impersonating Milla Jovovich (by Georges Biard / CC BY-SA / cropped from https://goo.gl/GlsWlC); (c) SB impersonating SC; and (d) SC impersonating Carson Daly (by Anthony Quintano / CC BY / cropped from goo.gl/VfnDct). Credit: Mahmood Sharif et al.

Glasses allows you to stare at people without getting caught. Additionally, it is the secret weapon for someone avoiding surveillance. The scientists from the University of North Carolina at Chapel Hill used a change in physical appearance to confuse face recognition software. They have developed an eyewear with a wild pattern that has the trick to avoid fool recognition. The patterns manipulate an image.

READ: Cillia- New Software Redefines 3-D Printing

It has two impressive gains- it does not only fools face recognition software but also appearing to look like other people. It detects the difference between one face and the other which they wanted it to be mistaken for. Scientists also get successful in working out how to confuse the software.

Scientists printed bespoke patterns onto the front of the frames. This allows the wearer to obscure their identity and to impersonate people who look completely different.

Scientists said, “When worn by the attacker whose image is supplied to a state-of-the-art face recognition algorithm, the eyeglasses allow her to evade being recognized or to impersonate another individual.”

Timothy Revell, explained, “A white male researcher wearing the glasses was able to pass for American actress Milla Jovovich while a South-Asian female colleague was digitally disguised as a Middle-Eastern male.”

The system focuses on things like the color of different pixels. After that, it slowly pieces together the best guess of who’s in the shot by comparing it to other, similar images. If just a small area of the face has been changed, it can completely mess with the attempted recognition. That’s the reason the computer system can confuse two people who in fact look very different.

Such kind of techniques generates accessories in the form of eyeglasses frames- when printed and worn, can effectively fool state-of-the-art face recognition.

One of the co-authors said, “We’re starting to find that neural networks don’t always have the flexibility that we once thought they had,” and just “a few small targeted changes can have a large overall effect in tricking the software.”

Revell said, “The frames may just look like a colorful design. But, The frames essentially overlay the face with pixels that perturb the software’s calculations in just the right way that it misidentifies the person as another specified face in its database.”