People Are Losing Their Minds Over FaceApp

A stock male model as seen through the FaceApp. Credit: Getty Images

There’s no shortage of apps out there for playfully manipulating and rearranging photos of our faces to make them older, younger, prettier, uglier, and so on. The latest addition to the club, however, has caught an especially ripe wave of hype.

People are going nuts for FaceApp recently. It’s an automated picture editor generating finished results that vary from awesome to terrifying. Feed it a photo of a face and it will let you adjust the person’s age and gender. It’s a lighthearted (and even gimmicky) application, but it packs some impressive technological cred under the hood.

FaceApp uses neural networks, and artificial intelligence make its facial modifications. Snapchat filters, for example, use basic facial recognition to identify parts of your face and display a filter over them. But FaceApp is powered by deep-learning algorithms that let it work its twisted magic on your photos. You never know what you’ll get as you twist FaceApp’s knobs, but that’s half the fun.

FaceApp CEO Yaroslav Goncharov touted his product’s technological superiority to TechCrunch in February, saying, “This technology is quite mature for some tasks, such as artistic style transfer or super-resolution, but has extreme challenges for photo-realistic tasks especially with high resolution images. I don’t think similar effects can be achieved with conventional (not deep learning) algorithms.”

The app’s controversial and now-defunct “Hot” filter generated heated discussion online. The filter removed blemishes while simultaneously lightening the skin tone. It moved many to shout racism.

FaceApp responded by killing the feature. Goncharov issued an apology to the Guardian, saying, “We are deeply sorry for this unquestionably serious issue. It is an unfortunate side-effect of the underlying neural network caused by the training set bias, not intended behavior.” In other words, the AI technology was following the skin color preferences it had learned in its human-sourced data set. As of April 25, it has been free of any racist filters. If you want to experiment with the latest breakout app, you can download it for free on Android or iOS.