Facebook learned to recognize people without a face

Facebook's face recognition technology has reached such a level that it is able to identify a person even if faces are not visible at all (for example, from the back). This was made possible thanks to the latest achievements of computer vision systems, writes New Scientist.

The experimental neural network developed in the AI ​​laboratory has been trained using additional features, such as hairstyle, hair color, clothing, body shape, and posture.

Conventional facial recognition systems already work quite well, they are equal to the person, and some even bypassed the person. Facebook decided to go ahead and try to teach the program to recognize people, even if a person is hard to see or not visible at all, because people do a good job with this task, explains the head of Artificial Intelligence on Facebook, Yann LeCun.
')
In fact, even looking from the back there are quite a few signs by which we recognize a familiar person. Let's say Mark Zuckerberg is easy to find out, because he always wears the same T-shirts.

The neural network was trained on 40,000 photos from Flickr, in some of which the face was clearly visible and in others it was bad. After training, the system showed recognition accuracy of 83%.

The results of the study were presented earlier this month at a conference on computer vision and image recognition in Boston.

In the future, the algorithm may be embedded in commercial products Facebook. For example, in the program Moments , which performs automatic sorting of the archive of photographs by individual events, with recognition of the names of all friends in the photographs.

So sunglasses and hoods will no longer hide you from Big Brother, %% username.