Interview with Paul Debevec, first recipient of the Significant New Researcher Award

(eMail interview)

What first drew you to computer graphics?

In grade school I was very interested in math and also loved to draw and paint. When I realized that with computer graphics I could combine these interests I got very intertested in programming computers to make pictures. I also picked up photography and filmmaking in high school, and thus became very interested in manipulating photographs with a computer, which led me to the particular kind of computer graphics I've worked most in, which is image-based modeling and rendering.

Do you have any favorite CG mentors?

The first person I ever worked for who completely amazed me was Michael Naimark, who was a media artist at Interval Research Corporation and supervised an internship I did there in 1994. He had equal respect for art and technology and used the two together to create amazing projects that greatly resonated with what I found interesting. I'm also very fortunate to have had the advice and encouragement of Pat Hanrahan, David Salesin, Gavin Miller, Paul Heckbert, Ned Greene, Don Greenberg, Michael Cohen, Lance Williams, Carlo Sequin, Brian Barsky, David Forsyth, Ken Musgrave, Jim Kajiya, Ken Perlin, and especially Greg Ward and Marc Levoy at important points along the way, and I'm also very appreciative of Ramesh Jain, Jitendra Malik, C.J. Taylor, and Quon-Tuong Luong for helping me leverage computer vision techniques in my computer graphics work.

What was the first time you contributed to SIGGRAPH?

It was 1996 in New Orleans, when I presented the paper based on my Ph.D. Thesis "Modeling and Rendering Architecture from Photographs." I also had also worked with a colleague Golan Levin on a piece in the art show, "Rouen Revisited", which used the techniques from the paper to create an interactive exploration of the Rouen Cathedral and Monet's related series of paintings. I really liked having a piece in both the technical session and in the art show, and since then have worked to do something creative as well as technical whenever it's been possible.

What year/city was your first SIGGRAPH? Which was most intense? Why?

My first was Orlando in 1994. I'd never been to an event of that magnitude before and was left completely reeling. When I saw the Electronic Theater it was probably the most exciting show I'd seen in my life - research and art and movies all coming together in one place.

My most intense was 2000 in New Orleans - I was presenting in two courses, chairing three sessions, delivering a paper, speaking in two panels, had a film in the Electronic Theater and a demo in the Creative Applications Laboratory. I have no idea how I made it through the week!

What contributions to SIGGRAPH are you most proud of?

I'm most proud of my films "The Campanile Movie" and "Fiat Lux", which I tried to make work as short films on their own independent of the techniques used to make them. The former used the photogrammetric modeling and image-based rendering techniques from my first SIGGRAPH paper, and the latter used the high dynamic range photography, image-based lighting, and inverse global illumination techniques from my next three. It's been rewarding as well that some of the students who worked with me on these projects have gone on to apply some of the same techniques in Hollywood films.

What's your favorite thing at this year or last year's SIGGRAPH?

This year I loved Stanford's "Rendering Translucent Materials" work as well as the magnetic dancing fluid. I also really loved some of the parties - particularly the smaller ones with good food and good people.

What near/intermediate developments in CG do you look forward to?

I'm looking forward to more extensive use of high dynamic range imagery, which will enable the techniques we've been involved with to be more accessible to people, as well as more realistic and easier-to-animate virtual humans. I would greatly appreciate a better head-mounted (or otherwise) immersive display device; so that the worlds that we're creating can actually be displayed as realistically as they are rendered. And I look forward to more communication and cross-fertilization between the visual effects and graphics research communities.