Facial recognition slipped into Google image search
Facial recognition slipped into Google image search
By Jacqui Cheng | Published: May 30, 2007 - 11:56AM CT
Google upped its stalker factor this week by adding face recognition abilities to its image search. While currently unofficial and unannounced, users can now search for images that only contain faces by appending a query string onto the end of a search URL. For example, a general image search for "Ars Technica" produces a variety of image results, but when appending "&imgtype=face" to the end of the URL, all new results contain photos of people.
The hidden feature was discovered by Google Blogoscoped, and there is currently no way to indicate that you only want to search for faces through the image search interface. However, both "&imgtype=face" and "&imgtype=news" trigger different search results than what is presented by default—the latter showing only images that are associated with news stories.
The technology appears to be the fruit of Google's 2006 acquisition of Neven Vision, a company that had developed techniques for facial recognition in photos. "Neven Vision comes to Google with deep technology and expertise around automatically extracting information from a photo. It could be as simple as detecting whether or not a photo contains a person, or, one day, as complex as recognizing people, places, and objects," wrote Picasa product manager Adrian Graham on the Official Google Blog last August.
Google is apparently taking that technology to heart by experimenting with facial recognition online. Even cooler (or creepier, as the case may be), one day Google's image search may be able to find faces of specific people based on image analysis/recognition alone instead of relying on the text associated with that image to identify the person in the photo. We can probably expect more search parameters to be added to the search in the future, too, such as different types of animals, different clothing items, and more. Until then, we're stuck experimenting with different search terms in hopes of discovering one that the public doesn't know about yet.
