Future Tense

Facebooked in the Crowd

Mobile face recognition technology means you can snap a photo of someone and pull up their online identity.

Photo by Karyn Poupee/AFP/Getty Images.

Chances are Facebook knows a lot about you. But can it pick you out in a crowd?

Probably not yet, but that could be on the horizon. Facebook on Monday confirmed that it is buying the five-year-old Israel-based startup Face.com, maker of facial recognition software. (AllThingsD put the pricetag at about $60 million.)

Facebook already uses Face.com’s technology in its photo tagging suggestions, which attempt to guess which of your friends are in pictures that you upload to the site—occasionally with awkward results. Now that it owns the company, the assumption is that Facebook will improve on the service and look to unveil facial recognition for pictures you take on your mobile phone, too.

From a business perspective, the acquisition seems natural: Facebook knows mobile photo-sharing is central to its future, and the Face.com buy dovetails with its purchase of Instagram and launch of Facebook Camera.

From a human perspective, though, the prospect of Facebook building up an ever-wider database that links people’s faces to their personal information is a little eerie.

In a study published last year, a team of Carnegie Mellon researchers found that they could combine off-the-shelf facial recognition software—in this case, Google’s PittPatt technology—with Facebook data and a computer algorithm to guess, not only people’s names, but in some cases their social security numbers, based solely on snapshots taken with a webcam.

This only worked for a minority of the photos in the study. People who didn’t have any public Facebook photos were mostly immune to identification, says Alessandro Acquisti, the study’s lead author (though at least one subject found that he had been tagged publicly in a friend’s photo without his knowledge.)

But facial-recognition software is improving rapidly. And software like Face.com’s gets better and learns more every time someone uses the tagging suggestions and clicks “yes” or “no” to indicate whether they were correct. “They’re being smart in a way, or some could say very subtle, in enlisting users as a means of improving the accuracy of their identification,” Acquisti told me.

As Facebook’s database develops, it’s conceivable that within a few years you could see someone on the street, point your iPhone at her, and pull up a list of possible identity matches within seconds. (In theory Google could do much the same, though unlike Mark Zuckerberg, Google’s Eric Schmidt has come out against the use of his company’s platform for facial recognition and mobile tracking.)

For now, Facebook only auto-suggests the identities of people who are among your friends. Still, the company will possess the information and capacity to identify and track people on a broad scale, both on the Web and out on the street—“much more than any government agency,” in Acquisti’s estimation. Only the company’s concern for your privacy will stand in the way.