“I’ve always been drawn to watercolour rendering and impressionism,” says Aaron Hertzmann, the 2010 recipient of the Steacie Prize for Natural Sciences. Looking at his body of work, there is no denying that Hertzmann has the eye and spirit of an artist. But he is not a painter — he’s a computer scientist. In fact, Hertzmann is only the second computer scientist in history to win the Steacie Prize, which recognizes outstanding research by Canadians 40 and under.
Hertzmann’s research is well-known for its influence in three separate areas of computer science: computer graphics, machine learning, and computer vision. In fact, Hertzmann’s work has resolved a wide range of problems in the field of computer graphics. These include estimating the three-dimensional structure of a non-rigid object from a video sequence of that object, finding new methods for removing the effects of “camera shake” from photographs in digital photography, computer rendering of images in diverse artistic styles, and automated construction of mathematical and computational models of human locomotion for computer animation in film and computer games.
Perhaps there is no finer example of the sheer artistry of Hertzmann’s work than in the 2009 independent animated film The Spine, in which he served as advisor to Academy Award-winning Director Chris Landreth. The film depicts a couple’s failing marriage, and uses an innovative computer animation technique known as non-photorealistic rendering. Whereas with photorealism artists paint images to look like photographs, non-photorealistic rendering strives to be as expressive as possible. Using smooth textures, loud colours, and fluid movements, the technique literalizes the characters’ internal conditions.
Indeed, much of Hertzmann’s work is focused on the very nature of art. In one of his papers, he asked, “How do artists create imagery?” and, “How do observers respond to artistic images?” These are questions Hertzmann hopes to answer — not as a philosopher, but as a computer scientist. His unique integration of computer science and art is rare, and could further pave the way for a new and emerging movement. Much of his research is also focused on designing “new and innovative tools for artists, designers, and scientists for creating beautiful images and animations.”
Hertzmann explains that he discovered the diversity of his interests at an early age. “I’ve always been interested in computers, and I’ve always been interested in art. I used to paint when I was little. Then I began creating art using computer algorithms.”
This interest would eventually take Hertzmann’s career to the animated film industry. He currently works at Pixar Animation Studios in Emmeryville, California as a visiting researcher. His work involves the response of digital characters to changing environments, and has potential applications not only for animated films but for the expanding video game industry.
However, his job as a Pixar researcher has not prevented him from performing the standard duties of a professor. Hertzmann also teaches several graduate and undergraduate courses at U of T, including a non-photorealistic rendering course and a computer graphics course.
Hertzmann is originally from Rice University in Houston, Texas, where he received his BA in computer science and art. He then moved on to his graduate studies at New York University before coming to U of T in 2003.
Explaining his initial reaction upon finding out he won the 2010 Steacie Prize, Hertzmann says he was “excited and happy.” Unfortunately, he did not find out about his success by way of a frantic phone call or an excited colleague, as one might expect — but a way perhaps most befitting a computer scientist: via email.
On where Hertzmann plans to take his research in the future, he explains, “There is still much work to be done on how physical characters interact with a variable environment.” Such research would have profound implications for the future of computer graphics and animation.
Hertzmann’s love of animated films is inspiring enough to recognize the present and future significance of computer-generated imagery. Computer-animated films have expanded rapidly since the mid 1990s, while the techniques of computer animation have since leaked into the methods of traditional filmmaking, forever impacting the cinematic tradition. Computer animation techniques have also evolved rapidly since the first use of CGI in the 1970s. Hertzmann’s work is undoubtedly a testament to this evolution. He strongly encourages viewers to take note of the cutting-edge advancements being made in animation by expanding one’s visual palette, and “viewing lower-budget non-mainstream animated films that demonstrate genuine innovation.”
Imparting some final wisdom to future computer scientists, Hertzmann states, “Do try to learn to love math. It’s extremely important and useful in the field.”
