Programmers Made a Selfie Filter To Make You Look More Like You, Sorry
For anyone who’s feeling selfie-conscious.
Image: Cat Faeries
Contrary to what internet culture says about selfies, they're hardly more than just photos of your face. Ideally, you'll look good in them. Possibly better than you look in real life.
But if there's one thing any selfie connoisseur understands, it's that, for many, using the front-facing camera on your phone can distort the heck out of your features. Your eyes are suddenly bulbous and alien. And from certain angles, you'd swear your chin could cut glass.
Knowing this, a team of computer scientists at Princeton University created a new method for making your selfies look more realistic—and, ultimately, more attractive. The editing software, which is the first of its kind, can modify your face as if it were professionally photographed, or even alter the height and angle of your portrait. The end result, according to its makers, is a lifelike 3D rendering, and not a flat, 2D image.
"Although it is the age of the selfie, many people are unaware of how much these self-portraits do not really look like the person being photographed because the camera is way too close," said Ohad Fried, the experiment's lead developer and a Ph.D. candidate at Princeton University, in a statement.
"Now that people can edit so many aspects of a photo right on their phones, we wanted to provide a quick way to edit faces that maintains realism."
The project began when engineers attempted to build a model for generating 3D heads. Using data from FaceWarehouse, the team was able to create six dozen reference points across the human face. When fed a selfie, the software will stretch and compress its subject matter, based on these anatomically correct coordinates. Then, voilà, a photograph that shows how people really see you.
"As humans, we have evolved to be very sensitive to subtle cues in other people's faces, so any artifacts or glitches in synthesized imagery tend to really jump out," said Adam Finkelstein, the paper's lead author and a professor of computer science at Princeton University.
As for when the tool will be made into a commercial product, researchers are still ironing out some technical kinks. For example, when repositioning your selfie, the synthesizer can't account for missing features, such as ears, if you're a fan of the egregious Myspace pose. And, occasionally, it has trouble interpreting hair.
I tested several photos of myself and Motherboard colleagues on the public beta version, but was told each time that "no face was found in the image." I suppose until it's fixed, we'll remain weird-looking and selfie-conscious.
- computer science
- motherboard show
- Princeton University
- facial reference
- front-facing camera