The new features takes advantage of Google’s machine learning abilities to make a set of stickers based on your face.
Feel like your feelings could better be expressed through emoji that look like you? Well, you’re in luck if you’re an Allo user. The latest update comes with a feature that turns your selfie into a reusable sticker.
The feature was announced in an official Google blog, and then confirmed in an interview between Google’s Jason Cornwall, Communications UX Lead at Google, and FastCo Design. Starting today, you’ll be able to shoot and save your own emotive stickers to use within Allo conversations. Google’s image-recognition algorithm will analyze your face and map each of your individual features to a preset selection of images illustrated by artist Lamar Abrams, who is best known for the Cartoon Network series Steven Universe. Then, you can choose the emotion that best suits your mood.
“The goal isn’t accuracy,” Cornwell said about the ability. “It’s to let someone create something that feels like themselves, to themselves.” Google estimates that there are 563 quadrillion possibilities. Once you create your stickers, you’ll have 22 different moods to choose from.
The FastCo Design article continues with some background on Google’s objective behind the sticker-making feature:
The project represents a long-running priority at Google—to figure out new ways that it can apply ML to broader and broader swathes of experience. The logic, for Google, is alluring: Google leads the world in ML, so if it can make ML into a must-have feature for apps and websites, then its products will be able to leapfrog competitors. Along those lines, Allo has become a test bed for all kinds of novel ML applications. “What we’re doing with Allo is trying to find all the ways that ML can make messaging better,” says Cornwell. “From saying the right thing at the right time to conveying the right emotion at the right time.”
Unfortunately, I don’t see the update available in the Google Play Store at the time of writing, which is quite a bummer. A sticker-making feature that uses Google’s Machine Learning abilities is rather impressive, and I’m curious to see the end result in real time.