Google shoots for visual search with “Google Goggles”
No, “Google Goggles” aren’t something off Urban Dictionary that you can drop into an e-mail when you’ve ended up in bed with a wealthy geek.
Google announced plans today for a new visual search, allowing searchers to submit queries via photograph, drawing on the powerful database of information behind Google Image Search. Vic Gundotra, Google’s VP of engineering, described the initiative:
“It is our goal to be able to identify any image,” he said. “It represents our earliest efforts in the field of computer vision. You can take a picture of an item, use that picture of whatever you take as the query.”
Gundotra fleshed out the merits of the visual search function with a dinner-party anecdote. Gifted with a bottle of wine he was clueless about, he used a visual search to quickly crib knowledge about it and properly express gratitude. During the event in which Google Goggles were announced, Gundotra also used them to correctly identify the Itsukushima Shrine in Japan- one of those things you’d know if you saw but not know the name of unless you’re some kind of trivia dork… clever.
Now, before you privacy humpers get all stressed out, facial recognition is not on the agenda for Google Goggles… yet. Scarily, the function works well to recognize faces, but Google’s wisely chosen not to allow it, which just made a whole bunch of girls on the C train breathe a bit more easily.
At the event, Google also demoed an English to Spanish translation using a cell phone.