The Next Generation of Emoji Will Be Based on Your Facial Expressions

A new app is trying to make it simpler to help you react to photos and videos that your friends post online—it’s using AI to capture your facial expressions and automatically translate them into a range of emoji faces.
Polygram, which is free and available only for the iPhone for now, is a social app that lets you share things like photos, videos, and messages. Unlike on, say, Facebook, though, where you have a small range of pre-set reactions to choose from beyond clicking a little thumbs-up icon, Polygram uses a neural network that runs locally on the phone to figure out if you’re smiling, frowning, bored, embarrassed, surprised, and more.
Marcin Kmiec, one of Polygram’s cofounders, says the app’s AI works by capturing your face with the front-facing camera on the phone and analyzing sequences of images as quickly as possible, rather than just looking at specific points on the face like your pupils and nose. This is done directly on the phone, using the iPhone’s graphics processing unit, he says.
When you look at a post in the app (for now the posts seem to consist of a suspicious amount of luxury vacation spots, fancy cars, and women in tight clothing), you see a small yellow emoji on the bottom of the display, its expression changing along with your real one. There’s a slight delay—20 milliseconds, which is just barely noticeable—between what you’re expressing on your face and what shows up in the app. The app records your response (or responses, if your expression changes a few times) in a little log of emoji on the side of the screen, along with those of others who’ve already looked at the same post.
The app is clearly meant to appeal to those who really care about how they’re perceived on social media: users can see a tally of the emoji reactions to each photo or video they post to the app, as well as details about who looked at the post, how long they looked at it, and where they’re located. This might be helpful for some mega-users, but could turn off those who are more wary about how their activity is tracked, even when it’s anonymized.
And, as many app makers know, it’s hard to succeed in social media; for every Instagram or Snapchat there are countless ones that fail to catch on. (Remember Secret? Or Path? Or Yik Yak? Or Google+?) Polygram’s founders say they’re concentrating on using the technology in their own app for now, but they also think it could be useful in other kinds of apps, like telemedicine, where it could be used to gauge a patient’s reaction to a doctor or nurse, for instance. Eventually, they say, they may release software tools that let other developers come up with their own applications for the technology.
Keep Reading
Most Popular

Meta has built a massive new language AI—and it’s giving it away for free
Facebook’s parent company is inviting researchers to pore over and pick apart the flaws in its version of GPT-3

The gene-edited pig heart given to a dying patient was infected with a pig virus
The first transplant of a genetically-modified pig heart into a human may have ended prematurely because of a well-known—and avoidable—risk.

Saudi Arabia plans to spend $1 billion a year discovering treatments to slow aging
The oil kingdom fears that its population is aging at an accelerated rate and hopes to test drugs to reverse the problem. First up might be the diabetes drug metformin.

Yann LeCun has a bold new vision for the future of AI
One of the godfathers of deep learning pulls together old ideas to sketch out a fresh path for AI, but raises as many questions as he answers.
Stay connected

Get the latest updates from
MIT Technology Review
Discover special offers, top stories, upcoming events, and more.