The Facebook project known as Deep Face can decern the accuracy of the true identity of any picture of you. DeepFace AI system is now powerful enough to spot users in the 400 million photos uploaded to the social network every single day.
Facebook is claiming good intentions with the Deep Face program. Facebook claims that instead of tagging users in embarrassing and incriminating photos without their permission, users will be able to first see the photos they are appearing in and then have the choice to blur out their faces.
The Deep Face program works for strangers’ pictures as well as ones from friends, but users can only see the identities of people they already know.
But with good or bad intentions, Facebook holds all the cards and will still be able to identify you in any picture and hold on to that very valuable personal data. Additionally, other entities who are researching similar technology specifically the government and private companies such as Google may very well have access to this data as well.
The technology behind Deep Face is actually very intriguing. With the ability to read features in a variety of lighting conditions and angles the way human eyes are able to, Deep Face uses an algorithmic technique called “Deep Learning”. By drawing on constants from an existing image data, the program is able to learn to recognize pixel patterns in new faces and become more accurate in identifying people’s faces. The algorithm improves itself, and is able to analyze faces as eyes, mouths, and ears instead of pixels and use that data to guess when the same faces show up in vastly different kinds of pictures..
At the end of the day, has its hands full with convincing people to believe it is not nefarious with its intentions with Deep Face. How they go about this will be interesting and should be watched carefully.