Earlier this year Google announced a new image detection feature they had been working on called Google Lens. While on the face of things it functions similarly to an older application of theirs, Google Goggles, the technical aspect had changed drastically so it evolved into a new service. We recently showed you how to try out Google Lens Launcher for Google Photos, but today the company has announced they’re launching it in a preview state for Pixel 2 customers.
At its core, Google Lens is a service that can be used to identify objects in a photo. The company is able to take things much further though by offering actionable things with these detected objects. For instance, at the reveal during Google I/O, the company showed that you can point your camera at WiFi network details (SSID and password) so the device can instantly connect to the network. At the reveal, they also showed it detecting a flower and a storefront while offering you details about what you’re pointing the camera at.
As Google has continued to work on the technology, they’ve been able to expand the type of objects and actions that Google Lens can handle. During today’s event, we learned that it can identify artwork, barcodes, books, buildings, landmarks, media covers, music albums, movie posters, places, points of interest, statues, video games and much more. This list of objects that can be detected has been achieved thanks to Google’s machine learning technology and thus it is able to grow without requiring the user to purchase new hardware.
As mentioned, Google Lens can also take action based on what you’re pointing the camera at. For some of the things listed above, this simply means doing a Google search for you, but like the WiFi network example, it can do much more. Today we also learned that Google Lens can add a contact from a business card, translate text, look up product information, open a web address in your browser, identify plants and animals and save dates to your calendar from a poster.
Excited for computer vision in personal assistants? What use cases are you interested in?
0 comments:
Post a Comment