Print 15 comment(s) - last by Ammohunt.. on Mar 20 at 6:26 PM

The software is called "DeepFace"

Facebook has created a piece of software that can tell whether two faces match with near-human accuracy. According to MIT Technology Review, the new Facebook software is called DeepFace, and it utilizes deep learning for facial verification.

Deep learning is a new branch of artificial intelligence (AI) that uses networks of simulated neurons to recognize patterns in large quantities of data. And facial verification is different from facial recognition, because it recognizes that two images reveal the same face where the latter puts a name to a face. 

What's more is that DeepFace can accomplish facial verification whether the person is facing forward or not. This is how it works: DeepFace uses a 3-D model of an “average” forward-looking face to adjust the angle of the face it's attempting to verify. This makes the face forward-looking. Then the simulated neural network configures a numerical description of the reoriented face, and if DeepFace finds descriptions from two different images that nearly match, it concludes they must be the same face.

DeepFace [SOURCE: MIT Technology Review]

In testing, the researchers put DeepFace up against a standard data set used to benchmark face-processing software.

When asked if two photos of faces are that of the same person, a human will answer correctly 97.53 percent of the time. DeepFace managed to get it right 97.25 percent of the time with the same challenges. 

The report added that changes in lighting will not affect DeepFace's accuracy. 

Facebook was able to make such a leap in AI because of its enormous database of photos. It was able to train DeepFace to recognize patterns with its copious amount of photos to choose from, where people are presented in many different poses in various environments within the photos. This made DeepFace more accurate over time.

DeepFace was made by Yaniv Taigman and other members of Facebook’s AI team. They will present their work at the IEEE Conference on Computer Vision and Pattern Recognition in June 2014. 

Source: MIT Technology Review

Comments     Threshold

This article is over a month old, voting and posting comments is disabled

RE: Please
By someguy123 on 3/18/2014 9:31:54 PM , Rating: 2
In a few years we'll actually be able to pan, zoom, enhance, spin, flip and then create a 3D model of someone else using an enhanced eye reflection. Then we'll realize that TV shows were actually ahead of the curve.

RE: Please
By BRB29 on 3/19/2014 7:38:18 AM , Rating: 2
TV shows, especially science fictions, gets their "ideas" from real ideas and theories.

"There is a single light of science, and to brighten it anywhere is to brighten it everywhere." -- Isaac Asimov

Copyright 2016 DailyTech LLC. - RSS Feed | Advertise | About Us | Ethics | FAQ | Terms, Conditions & Privacy Information | Kristopher Kubicki