More Than Meets the Eye

Kevin Jing at the Pinterest office.

A picture isn’t just worth 1,000 words to Kevin Jing. A picture is a source of inspiration, enrichment and a connection to other similar images scattered across the Internet. That connection is worth more than any amount of words -- and Jing has devoted a large part of his career to understanding them.

Jing, a triple College of Computing alumnus, is a visual search leader in the tech community. A Google veteran, Jing currently leads visual discovery efforts at Pinterest, after the start-up he co-founded was acquired in 2013.

It all started at Georgia Tech. Jing’s arrival in 1999 coincided with the height of the dot-com boom. He became interested in computer vision after taking the artificial intelligence class taught by Professor Thad Starner, and subsequently worked on an undergraduate research project with Professor James Rehg. Their research partnership would continue for seven years.

Jing (left) and Google colleagues in 2004

He continued onto a Georgia Tech Ph.D. program in computer science in 2003. During the second year of his study, he secured an internship at Google Research and helped to develop one of Google’s first image-analysis applications. He joined Google Research full-time in 2006, while simultaneously completing his Ph.D. in Atlanta, advised by Rehg and Google's Michele Covell

Fittingly, Jing chose to concentrate his professional endeavors on visual search, an ideal application of his computer vision and machine learning education.

“I initially investigated how to apply PageRank to web images,” Jing said. “We quickly discovered very interesting properties of image graphs inferred from visual similarity.”

His work resulted in more than 10 publications and 40 patents in the area of media search and advertising. In 2012, in order to make large-scale image-analysis tools available to other businesses, he left Google and co-founded the aptly named startup, VisualGraph, with the help of David Liu and two other College of Computing alumni, David Tsai (MS CS '12) and Michael Feng (MS CS '13).

“We built up a cloud-based, highly scalable visual-search platform that leverages the latest computer vision/machine learning techniques for product recommendation,” Jing said. “The idea was to help e-commerce and photo-sharing sites to better organize and monetize their contents.”

VisualGraph’s efforts caught the eye of Pinterest, which acquired the company in December 2013. The decision to accept Pinterest’s offer was quite easy for Jing.

Jing (front right) with Pinterest engineering colleagues in 2014

“It was a win/win situation -- Pinterest is rapidly becoming a human-powered visual-discovery engine, and we brought with us one of the best visual-search platforms to supplement that experience,” he said. “It was like joining Google in 2004 -- lots of very open-ended problems to define and solve.”

Take a snapshot of Jing’s personal life and you’ll find familiar passions.  He enjoys family, food, and watching college football. Jing was married last year and plans to start a family soon. He’s also a big fan of Yellow Jacket football, a passion that began as an undergrad.

At age 34, Jing has already accomplished so much, both personally and professionally. But the truth is, he’s just begun.

“I’ll be continuing the vision of VisualGraph at Pinterest, to connect all the visual information in the world together and make it useful for others,” Jing says.

Get the picture?