By Brittany Thompson
Modern technology appears to be taking the worldâ€™s leading search engine in a new direction that might allow Google to recognize facial features in images. Google Blogoscoped reported Monday that Google Images has added a few new features, which are only available by appending URL variables of current image searches. Such variables allow searchers to narrow searches to images of faces (people) or news-related images.
For example, a Google image search for the word â€œblogâ€ brings up various logos, graphs, and images from blogs around the world. Adding the variable &imgtype=face to the end of the results page URL, however, brings up only images of people â€“ many of whom are the actual bloggers themselves. Changing that end variable to &imgtype=news shows images from recent news features in the New York Times or the Washington Post, to name a few.
Google already examines file size and resolution to detect duplicate images. Additionally, the company openly encourages webmasters to use alt and title tags to accurately describe images for better search results. These new changes are apparently powered by Neven Vision, a facial recognition company that was purchased by Google back in August.
While completely visual image searches are probably still be a few years down the road, it does appear that Google is headed in that direction.