(TECH UPDATE) : This New Russian app lets you take photos of strangers and identify them online.


It sounds just like the premise of a dystopian sci-fi film, but people in Russia are going nuts for an app that permits them to take a photo of a stranger on the avenue, add it to an app, and figure out their identification, with around 70 percentage reliability. 

If that’s making you think mighty nervous, it should - imagine sitting on a educate and the guy opposite you is checking out your LinkedIn profile. The fellow at the back of you has already found your fb page and is now cyberstalking your satisfactory buddy. The lady subsequent to you has observed you on Tinder.
Creepy for a entire exceptional cause is the way the app can be used in reverse - add a snapshot of your superstar crush, and it'll bring up a bunch of strangers local who look like them.

Meet FindFace, a face recognition app developed by means of Russian 20-somethings Artem Kukharenko and Alexander Kabakov.

As ComputerWorld reports, the pair just lately "upstaged" over one hundred other facial awareness algorithms - together with Google's - within the school of Washington’s MegaFace assignment by means of achieving a seventy three.3 percentage accuracy on 1 million faces.

Considering it launched two months in the past in Russia, Findface has already racked up 500,000 customers and processed virtually three million searches," in keeping with Shaun Walker at The Guardian. It syncs up with Vkontakte - a standard social network in Russia and the previous Soviet Union, which boasts more than 200 million bills and just about 1 billion photos.

"Three million searches in a database of just about 1 billion pictures: that’s 1000s of trillions of comparisons, and all on four typical servers," Kabakov told Walker. "With this algorithm, you can search by way of one billion pictures in less than a second from a traditional laptop."

truly, all you have to do is sneakily image any person on the avenue, add the pic, and the app with serve up what it thinks is the certainly suit, and 10 'bonus' identities that appear an identical.
"in the event you see anybody you like, which you could graphic them, to find their identification, and then ship them a pal request," says Kabakov. "It also looks for equivalent men and women. So that you would simply upload a image of a film famous person you love, or your ex, and then in finding 10 girls who seem similar to her and send them messages."

*Throws laptop out the window, units fire to condo, moves in with a household of beavers*

if you consider creepy as hell courting systems are the worst of it… nope. The Guardian reports that individuals have already used the app to find individual social media bills of porn stars and harass them, and a photographer in St Petersburg has been taking snap shots of strangers in the metro and publishing them online with their identities - no permission required.  

"One lady within the task texted me after the newsletter and mentioned that it used to be a foul feeling when she noticed herself … but she completely understood my suggestion," photographer Yegor Tsvetkov informed Elena Cresci at the Guardian final month.

The pair behind FindFace insist that this is not the course they fairly need their program to head in. The purpose is for their program to help remedy crimes by with the aid of opting for criminals on the avenue, and make them a crapload of cash in the retail sector.

"Kabakov imagines a global where cameras fix you watching at, say, a stereo in a shop, the retailer finds your identity, and then pursuits you with advertising for stereos in the subsequent days," says Walker.

*sets fireplace to beaver house, strikes to Mars*

happily, the app only works with Vkontakte, and anything would need to dramatically exchange in facebook’s information storage methods for it to be ready to access any snap shots - even public ones - on there. So we’re reliable for now. However the notion and the technology is available in the market now, it pains me to assert it, but it’s simplest a matter of time.‎