Normal facial recognition software that we have embedded in our phones for unlocking or in apps like Facebook need to recognize our faces or else it fails. This is not the case with an experimental algorithm Facebook is working on in its Artificial Intelligence Lab. This new facial recognition software as reported by the News Scientist can recognize you even when it can’t see your face. Creepy has a face literally people
Apparently this algorithm also looks for other characteristics rather than your face like pose, clothing, and hairdo or body shape. You can understand why the social media company needed to develop this because normal facial recognition fails when it can’t recognize a face because maybe the photo is blurry or something.
This new creepy facial recognition software was apparently given a test by the research team who gave it 40,000 photos from Flickr for it to do its magic. The algorithm was reported to recognize individuals with about 83% accuracy. That is impressive!
Recently, Facebook launched Moments, an app that was specifically geared for people to organize and share photos you took with your friends easily between each other. The company said they’ll use facial recognition to determine which friends are in them. They have been using facial recognition for some time now but with the advent of the new algorithm, the company hopes to improve the accuracy so as to tag people accurately.
They say privacy is dead, and this new algorithm will clearly propagate that notion. Even if Facebook would want to use it for improving accuracy, the potential of it being misused by people to commit crimes is highly likely.
[…] isn’t the only one doing this malpractice. There’s a lot of collections of people’s photos done by academics who fail to point out where […]
[…] isn’t the only one doing this malpractice. There’s a lot of collections of people’s photos done by academics who fail to point out where […]
[…] recognition tech has come under fire in recent years since tech companies have been accessing giant databases of photos to improve the accuracy of their […]
[…] recognition tech has come under fire in recent years since tech companies have been accessing giant databases of photos to improve the accuracy of their […]
Comments are closed.