Subscribe / Unsubscribe Enewsletters | Login | Register

Pencil Banner

Facial recognition: Facebook photo matching just the start

John P. Mello Jr. | Sept. 21, 2011
The Internet was in an uproar earlier this year following Facebook's launch of facial recognition software for its photo services, enabling users to identify their friends in photos automatically--and without their permission. Though critics described that move as creepy, the controversial technology may now be on the verge of widespread use.

Though the members of London Riots Facial Recognition undoubtedly believed that they were working for the greater good, what happens when people other than concerned citizens get their hands on the technology? It shouldn't take too long for us to find out.

Present-Day Reality Check

The use of facial recognition software by governments and online social networks continues to provide headline fodder. A Boston-area man had his driver's license revoked because when the U.S. Department of Homeland Security ran a facial recognition scan of a database containing the photos of Massachusetts drivers, it flagged the man's license as a possible phony. Afterward it emerged that the system had confused the man's face with someone else's.

In England, law enforcement officials ran photos of August riot suspects through Scotland Yard's newly updated face-matching program, which is under consideration for use during the 2012 Summer Olympics in the UK. In Canada, an insurance company invited Vancouver police to use its facial recognition software to help identify rioting fans after the Vancouver Canucks hockey team lost the seventh game of the NHL championship series.

And of course Facebook endured a hailstorm of criticism in June when it announced its plans be roll out a facial recognition feature for its members to provide semiautomatic tagging of photos uploaded to the social network.

One Facebook critic was Eric Schmidt, executive chairman of Google, who said earlier this year that the "surprising accuracy" of existing facial recognition software was "very concerning" to his company and that Google was "unlikely" to build a facial-recognition search system in the future.

Indeed, Google seems to have been so concerned by the technology that Schmidt declined to implement it even though his company already had the know-how to make it. "We built that technology and withheld it," Schmidt said. "People could use it in a very bad way."

Next: Off-the-Shelf Efforts, Watch Out for Little Brother, and more

Off-the-Shelf Facial Recognition

You don't need the power of a government or an Internet behemoth to make facial recognition work for you. At this year's Black Hat security conference (held in Las Vegas in August), a team of researchers from Carnegie Mellon University demonstrated how much they could accomplish with existing off-the-shelf technology.

The team took photos of people's faces and pushed those images through an off-the-shelf facial recognition program called PittPatt (which Google recently acquired). In the demonstration, in less than 3 seconds, the program compared the CMU researchers' photos to images publicly available on Facebook and returned 10 possible matches, along with the names of the matches. The process proved to be accurate more than 30 percent of the time.


Previous Page  1  2  3  4  Next Page 

Sign up for Computerworld eNewsletters.