Facebook will recognize you in photos where your face is not showing

Facebook’s facial recognition blows me away every single time I see it at work. Some of us love it, as it makes it an easy task to tag all your friends and get on with your life. On the other hand, some people hate that Facebook knows so much about them. Well, get your tin foil hats ready, guys, because things are about to get a lot more interesting. Soon Facebook will no longer need to see your face in order to recognize you in images uploaded to their servers.

Facebook’s artificial intelligence lab has put together an algorithm that can scan a photograph and look for many other physical clues, hence being able to know who you are even when you are not showing your facial features.

Facebook tested this with almost 40,000 images taken from Flickr. These photos showed some people’s faces, but many other subjects had their faces turned away. This algorithm is impressive, to say the least. The system was able to recognize people with an 83% accuracy, which is simply stunning.

It makes complete sense for Facebook to improve their recognition technology in this manner. After all, it’s something us humans do all the time. I can recognize most of my friends even when they have their backs completely turned to me. The algorithm will probably take into account body features, hair, clothing and other factors.

The social network is also currently pushing Facebook Moments into the market. This photo-sharing service revolves around Facebook’s power to recognize you and your friends. The app scans your phone for photos and allows you to easily share them with your friends (those the app recognizes in your images). It’s Facebook’s easy way to take care of image sharing, and odds are it will also adopt Facebook’s new algorithm at some point.

I can’t deny this all sounds a little freaky, but I am not really trying to hide from the internet. This will offer very convenient features for Facebook users. What do you think, though? Are you comfortable with Facebook implementing such extreme identity recognition practices? Let us know in the comments below!