“Much of my research lies at the interface of computer graphics and computer vision,” says Ravi Ramamoorthi, the senior faculty hire this year for the Computer Science and Engineering department at the University of California, San Diego. “I have published and been active in both of these communities, so I have tried to build bridges between them.”
In addition to building bridges, effective July 1, Ramamoorthi has embarked on a new mission in San Diego. “My goal is to build visual computing at UC San Diego – graphics and vision – into the best group in the world,” he says matter-of-factly. “My hire is the first step in this process. I believe we now have the nucleus to make the leap as we add more faculty members. We won’t be #1 next year, but we are hopefully in a strong upward trajectory.”
CSE department chair Rajesh Gupta agrees. “We are actively looking to hire at all levels: junior and senior faculty, postdocs and graduate students,” says Gupta. “We brought Professor Ramamoorthi on board because he is a world-renowned expert in the field of computer graphics, and with CSE in a fast growth period, he can help propel UC San Diego to the top in both size and quality of research product.”
The newcomer and his colleagues – CSE professors David Kriegman (in computer vision) and Henrik Wann Jensen (in computer graphics) – expect to add at least two more faculty members, and hopefully more. Also ahead, they intend to build a joint research center and to go after large-scale NSF and industry support, while also beefing up the range of courses available to students. “We want to have a full teaching program in graphics and vision to expose students to the full range of the field,” says Ramamoorthi.
For his own part, Ramamoorthi will begin teaching a graduate course this winter at the intersection of graphics, vision and photography. Then next spring, he will launch a new Advanced Graphics course for upper-division undergraduates. He also intends to teach an online course for a broad audience later this year. His prior online course, “Foundations of Computer Graphics”, last ran on the EdX network in Fall 2013, and is also available on the UC Berkeley website.
When Ramamoorthi joined the Columbia University faculty in 2002, Columbia was just starting to build its Vision and Graphics Center. “We went from zero papers in SIGGRAPH four years earlier to having ten papers four years later – more than any other group that year,” he recalls. Then in January 2009, a young family in tow, Ramamoorthi moved back to California. He joined UC Berkeley to help rebuild their program in computer graphics. “There wasn’t that much activity in graphics at Berkeley the previous few years, but we built that into what you can argue is the best graphics group in the country,” he notes. “Building these groups has been an exciting part of my career so far, and I hope to do an even better job at UCSD.”
Asked why there is growing interest in visual computing as a field, Ramamoorthi notes that more than 90 percent of the sensory information we receive is through our eyes. “Computer graphics and computer vision are becoming fundamental and intrinsic to our everyday lives,” he argues. “We’ve already seen the impact on movies and gaming, but increasingly you’ll see it on your mobile phone, which now has a graphics chip as powerful as you have in your desktop computer.”
Ramamoorthi did his undergraduate work at Caltech, where he completed a B.S. and dual M.S. degrees in computer science and physics in 1998. He then earned his Ph.D. in computer science at Stanford University in 2002 with a dissertation on spherical harmonic lighting (now widely used in videogames and movies – including for the production of “Avatar” and various animated films from Pixar, where he was a consultant for three years). While working in Stanford’s Computer Graphics Laboratory, he established a firm mathematical framework to describe reflection – leading to a deeper theoretical understanding of light transport and its impact on the visual appearance of objects.
The graphics guru has collaborated with UC San Diego faculty and students for more than a decade. He and Henrik Jensen worked in the same lab at Stanford (under Pat Hanrahan) and completed their Ph.D./postdoc the same year (2002). Ramamoorthi has co-authored papers with Jensen and former CSE Prof. Serge Belongie (who moved in January to Cornell NYC Tech), and with CSE alumnus Sameer Agarwal (Ph.D. ’06), among others. For two years, his postdoctoral researcher at Berkeley was Manmohan Chandraker, a CSE alumnus (Ph.D. ’09), whose advisor was CSE Prof. David Kriegman. Earlier, while at Columbia, Ramamoorthi hired a postdoc, Craig Donner, who had been a UCSD graduate student (Ph.D. ’06) and then a postdoc, both under CSE’s Jensen.
In 2011 Ramamoorthi received a prestigious award from Japan’s Okawa Foundation for information and telecommunications. The foundation funded his work on “a digital visual appearance pipeline in computer graphics.” Earlier, he and 66 of the most “promising researchers in the nation” received the 2008 Presidential Early Career Awards for Scientists and Engineers from President George W. Bush. (Ramamoorthi was nominated, based on his work in physics-based computer vision, for the award by the Department of Defense.) A year earlier ACM SIGGRAPH gave him its Significant New Researcher Award, citing his “seminal contributions to the visual appearance of objects.” Ramamoorthi picked up his award at SIGGRAPH 2007 in San Diego. In its news release about the award, ACM lauded the then-Columbia University professor for developing “mathematical and computational models that have led to a deeper understanding of visual representation by digitally recreating or rendering complex natural appearances.”
Ramamoorthi was also one of only two computer scientists to receive a Young Investigator Award from the Office of Naval Research (ONR) in 2007, and recently received another ONR grant. Over the past 20 years, Ramamoorthi has attended all but one SIGGRAPH conference – the premier venue in computer graphics – and his Berkeley team will have four papers at SIGGRAPH 2014, which runs August 7-13 in Vancouver.