Google’s PlaNet beats humans in guessing where photos were shot

c/o Google Google’s new PlaNet deep learning network is capable of determining where a photo was taken, simply by looking at the image and analyzing it. And, per recent tests, it happens to be better than humans, even those who have traveled, or still travel quite a lot. The PlaNet network was designed by the tech giant to recognize certain locations on a photo, with a database of over 90 million geotagged photos used to figure out where it was taken. That could mean, for example, the system recognizing that the photo was taken in Las Vegas because it remembers the Mandalay Bay Casino, or in San Francisco because it recognizes the Golden Gate Bridge. But that’s not all Google’s new system can do. While humans are able to do the…


Link to Full Article: Google’s PlaNet beats humans in guessing where photos were shot