Google shared a lot of news at its Google I/O developer conference this week, ranging from a new version of Android that emphasizes digital wellbeing and new Android TV devices to a relaunch of Google News and artificial intelligence passing as humans in phone conversations.
But the biggest story of the event was Google’s advances in the field of computer vision and augmented reality.
Those advances flew a bit more below the radar, perhaps because they were scattered across a number of announcements and initiatives. But taken together, they showed that Google is getting ready to augment everything, and bring layers of information, games, and more to the space around us — all with the help of our phone cameras.
One of those examples: Google has been looking to bring computer-vision-based navigation to Google Maps. In the near future, consumers could be able to open the camera view of their phone from within Maps, point it at a street corner, and get instant directions on where to go. “Our teams have been working really hard to combine the power of the camera, the computer vision, with Street View and Maps to reimagine walking navigation,” said Google Lens vice president Aparna Chennapragada Tuesday.
Google is also expanding Lens, a feature it first introduced at Google I/O 2017, which adds computer vision smarts to the camera to identify dogs, buildings, plants, and other objects. After previously integrating Lens into Google Photos, the company is now bringing this feature directly to the camera app on a number of Android phones.
Thanks to new features being added to Lens, users of these phones will be able to just point their camera at documents and instantly select text any printed text. “You can do things like copy and paste from the real world directly into your phone,” said Chennapragada. Lens will also get better at identifying distinct objects, like books and posters.
And over time, Google even wants to overlay pieces of information directly over those identified objects – including music videos over concert posters. “This is an example of how the camera is not just answering questions, but it is putting the answers right where the questions are,” she said.
Google is also making its Google Photos service smarter, giving it the ability to automatically recognize when you take the picture of a document, and then convert it into a PDF file — something that got cheers from the audience Tuesday.
And finally, Google is expanding its efforts in the mobile augmented reality space. Google unveiled ARCore as its answer to ARKit last summer. The AR developer framework was at first limited to select Google devices and apps, but the company started to make it available to many millions of Android phones, as well as the wider developer community, in February.
On Wednesday, Google engineering manager Eitan Marder-Eppstein outlined a number of new additions to ARCore, including the ability for developers to build multiplayer AR experiences that can be shared across devices. This makes it possible to, for instance, play an AR Tic-Tac-Toe game, with one player using an iPhone and another using an Android phone.
Google’s AR technology also now allows developers to utilize augmented images. This could enable consumers to hold their camera up to a book cover or another physical product, and then have characters step out of that product and march around on your coffee table. “Imagine a textbook coming to life in front of you,” said Marder-Eppstein.
Another obvious application for these AR images is advertising, which could be a big draw for magazine publishers and other media companies. Said Marder-Eppstein: “Advertising is all about engagements.” And what’s more engaging than ads coming to life in the middle of your living room?
All of these announcements may have been disjointed examples of what Google is doing in the computer vision space, but they were tied together by combining cellphone cameras with artificial intelligence, and they laid a blueprint for a world in which visual objects offer cues to whole new layers of information.
It’s a change that is very profound, even if AR games and ads may seem playful and silly at first. Said Chennapragada: “Vision is a fundamental shift in computing for us.” It’s a shift that is not lost on other tech companies, which is why Facebook, Apple, and Snap are all busy exploring augmented reality and similar technologies as well.
But this week, Google outlined that it has a very far-reaching goal for this space: The company wants to augment everything, and turn your phone camera into a tool to browse the world.