Invisible, targeted infrared light can fool facial recognition software into thinking anyone is anyone else


#1

Originally published at: https://boingboing.net/2018/03/26/the-threaten-from-infrared.html


#2

from wikki about the character:
" The Laughing Man. He is an expert hacker, able to hide his physical presence by editing himself out of video feeds and cybernetic eyes, concealing his identity by superimposing an animated logo over his face, and hijacking cybernetic brains altogether, all in real-time."


#3

image


#4

I just re-watched all of GITS over the weekend.


#5

I’ve lost count…is this the third or fourth nail
in the iPhone X coffin?


#6

Normally I get concerned about security work-arounds, but anything that breaks facial recognition has a big tick from me.


#7

Thanks a lot you damn researchers. Now they’ll ban hats on airplanes. (and in airports)


#8

They can’t ban glasses. And LEDs and batteries can be small.


#9

I was thinking more A Scanner Darkly.


#10

feature-right
I prefer this to the last facial recognition spoof. Ballcap just gets less attention than the Bladerunner look in the South.


#11

I don’t think this particular attack would give false recognition with iPhone X because the Face ID system uses a 3D “depth map” and not only the reflected light from the skin surface but also below-surface aspects of the face.


#12

The thing that is most crazy about this is how one of the least believable aspects of 80’s cyberpunk turned out to be true. Actual cyberpunks would in fact have to dress and wear makeup like cubist circus clowns. If it turns out that mohawks are especially good at defeating computer based person identification I’m going to ask William Gibson for some lotto numbers.


#13

No thanks, I prefer my scramble suit. scramble-suit


#14

Does morris dancing still work?


#15

Sweet. Especially that they used The Man Who Invented the iPhone. I approve.


#16

I find this type of hacking to be interesting especially since the Victims photos doesn’t look anything like the attackers photos. I’m curious on what kind of face recognition software was used for this test. The paper states that the gadget used focus on the blind spot in the learning machine models. Is this software used a creditable face recognition or one that someone just created in their basement. I agree that face recognition is in its infancy state and needs to look at all potential threats before using for secured authentication purposes. The one concern I have with using face recognition is that if the images are stored in a database digitally somewhere what happens if it is breached? What access who the hackers have?
Also there are articles that claims that face recognition has been proven to be safer than fingerprints and voice recognition. Is one safer than the other or do all of them have their own vulnerabilities?


#17

This topic was automatically closed after 5 days. New replies are no longer allowed.