Facebook researchers have introduced a new artificial intelligence system that recognizes faces almost as accurately as a human does.
The "DeepFace" system identifies a person correctly from two photos 97.25 percent of the time, which is extremely close to a human's 97.53 percent accuracy, CNNMoney reported. The software's artificial intelligence imitates the human brain's structure to make connections within large sets of data.
How does DeepFace differ from the controversial face recognition technology that Facebook already uses? The system mimics the structure of the human brain and how neurons make connections, bringing it eerily close to people's capability for recognition.
The new software, which Facebook researchers detailed in a recent study, can view a photo and then create a 3-D model of a person's face using something called "Deep Learning," an artificial intelligence technology that can analyze an image based on more than 120 million parameters.
Facebook users don't need to worry yet; the social network hasn't put the technology to use.
The recent study was "theoretical research," said Facebook spokeswoman Lydia Chan, as quoted by CNNMoney. "[W]e don't currently use the techniques discussed in the paper on Facebook," she said.
Facial recognition technology is moving forward more quickly than any guidelines for the innovative software. While the Federal Trade Commission issued recommendations for such technology in 2012, privacy advocates have said the guidelines weren't enough.
The American Civil Liberties Union said the FTC recommendations avoided the true issue, which is whether face recognition technology and video surveillance will end public anonymity, CNNMoney reported.
In the new study, the Facebook researchers acknowledged that society will have to contend with questions of surveillance and anonymity. DeepFace has essentially eliminated the distance between technology and humans when it comes to facial recognition, they said.
"The social and cultural implications of face recognition technologies are far reaching, yet the current performance gap in this domain between machines and the human visual system serves as a buffer from having to deal with these implications," the researchers wrote, as quoted by CNNMoney.
They went on to write that the DeepFace system "has closed the majority of the remaining gap."
See Now: OnePlus 6: How Different Will It Be From OnePlus 5?
