Researchers for Google recently presented a paper on applying the company’s PageRank technology to the world of product image search.

Without appended metadata, search engines can’t truly see images. Plenty of work continues in the field, as unlocking a truly effective image search should be a profitable discovery.

On the official Google Research blog, presenting researchers Shumeet Baluja and Yushi Jing discussed some of the theory behind their paper, “PageRank for Product Image Search.” The main point: you can’t play with it yet, but if you’re qualified Google might hire you to help build it.

Popularity of product searches made that an ideal candidate for research into image search, the researchers noted. “Since the publication of the paper, we’ve also extended our results to other query types, including travel-related queries,” they said.

As they mentioned in April, VisualRank poses a massive computational demand. Right now, there isn’t going to be a massive all-encompassing index of images.

The researchers hope to build relevant subsets of product images, and use these to drive the results people see. In this latest post, they hinted “we have an approach that has an easy integration with both text and visual clues.” More on that in coming months.