Where on earth was that random photo taken? Google\u2019s freaky new artificial intelligence machine can figure it out.<\/p>\n
Computer vision specialist Tobias Weyand and his colleagues at Google created a deep-learning program called PlaNet, and trained it to identify locations where photos were taken based on visual cues.<\/p>\n
Putting the Art in Artificial Intelligence: Photos<\/a><\/p>\n Imagine the ultimate game of \u201cWhere in the World is Carmen Sandiego?\u201d Only way harder. The Googlers started\u00a0by dividing up the globe into a grid, excluding the oceans and polar regions. Then they created a database for PlaNet that contained 126 million geolocated photos pulled from the Internet,\u00a0Technology Review<\/a>\u00a0reported.<\/p>\n Since PlaNet is an artificial neural network, it can learn. So the team taught the network how to identify a photograph\u2019s location on the grid just using information contained in the pixels.<\/p>\n To test PlaNet\u2019s accuracy,\u00a0Weyand and his team fed it 2.3 million geotagged Flickr images. From there, PlaNet narrowed down 48 percent of them to the right continent, 28.4 percent to the right country, 10.1 percent to the right city, and\u00a03.6 percent to the actual street.<\/p>\n OK, so maybe it can\u2019t accurately locate every single random image on a map, but consider everything in that Flickr mixed bag: building interiors, pets, food. And while the results might not seem all that great at first, they became remarkable when the Google team pitted their machine against 10 smart, well-traveled humans.<\/p>\n The machine won more than half the rounds \u2014 and had better accuracy.<\/p>\n \u201cPlaNet outperforms previous approaches and even attains superhuman levels of accuracy in some cases,\u201d the team wrote in their abstract<\/a> about the machine.\u00a0You can test your own abilities with online games like\u00a0GeoGuessr<\/a>.\u00a0Might want to set an alarm first, though, because that one is kind of addictive.<\/p>\n A.I. Takes a Stroll Through Amsterdam<\/a><\/p>\n The team says that PlaNet doesn\u2019t need much memory, either. Their model only uses 377 MB, which means it could go into a smartphone, Technology Review reported.<\/p>\n