Google is already very successful due to its PageRank technology for web pages, and now the company has taken a further step by unveiling a cutting-edge image ranking technology called VisualRank. At the International World Wide Web Conference in Beijing, two Google scientists presented an algorithm which manages to perform image recognition and comparison of similar images to determine a ranking which will push to the first search results those pictures which are considered most relevant to a specific search.

Currently, images are ranked by analyzing the text near the image and the image's file name. “We wanted to incorporate all of the stuff that is happening in computer vision and put it in a Web framework,” said Shumeet Baluja, a senior staff researcher at Google. He was joined in presenting the new technology by fellow Google expert Yushi Jing.

Also, the company posted yesterday a blog entry summarizing the paper. Google's Shumeet Baluja and Yushi Jing explained that there are three directions that the VisualRank team is currently exploring: estimating similarity measures for all of the images on the web is way too costly, which means that approximate or alternative computations are needed; the large number of recently proposed alternative clustering methods are being considered in implementing VisualRank; and many variations of PageRank can be used to reintroduce, in a meaningful manner, the textual information which the initial Google Image Search used.

The new search method, VisualRank, cannot be used in the same manner in which traditional searches are performed. Analyzing a tremendous amount of images in depth requires, quite logically, incredible computing power. To overcome this problem Google has decided for now to perform analysis of the 2,000 most popular product queries on Google’s product search.

The paper presented by the two Google experts is called “PageRank for Product Image Search.”