Professor Shlomi Steinberg has a PhD in computer science from the University of California, Santa Barbara. While pursuing his doctoral degree he was a recipient of an NVIDIA PhD fellowship. He received his MSc in mathematics and computer science from the Weizmann Institute of Science in Israel under the supervision of Professor David Harel. His master’s research centred on efficient execution and distribution of formally verifiable software paradigms. Previously, he was a rendering engineer working on CRYENGINE, a leading game- and rendering-engine, and on HUNT: Showdown, a successful AAA video game.
Professor Steinberg’s research interests are in ray and wave optics, light transport and rendering. His research resides between the realms of computer graphics and computational optics. He is working toward developing a cohesive computational theory to simulate the wave-optical behaviour of light in complex, real-life scenes and environments. One application is simulating how optical and radar sensors in driving-assistive technologies perceive the environment. Accurate simulation of the car’s sensors would allow researchers to generate data for testing, validation, and training in a safe, virtual environment.
What follows is a lightly edited transcript of a Q&A interview.
When did you become interested in computer graphics and computational optics?
Computer graphics, rendering and optics are areas of research I’ve been interested in for a long time. I’ve spent many years, both in the industry and in academia, working in these areas.
If I step back a bit, it’s really about light. I’ve long been fascinated by how light works, how it interacts, how it behaves — the physics of light — and simulating that physics precisely and accurately on a computer, which is difficult to do.
Initially, I was interested in particular optical effects that you can observe and how these effects can be simulated on a computer, but my research interests now encompass a bigger picture. I’m interested in light behaviour as a global physical process. When I say light I don’t mean just the light we can see — what’s known as the visible spectrum — but also light we cannot see, the other parts of the electromagnetic spectrum used, for example, in radar and in cellular communications. Radio is a form of electromagnetic radiation just like light, and understanding and simulating such nonvisible bands on a computer is not only very important but also an open problem.
I also work on rendering, a part of my research that has applications in computer-generated imagery and computer games. Efficient wave-optical rendering is a side interest. It’s not as rigorous as simulating real light physics as we don’t need ultra-accurate simulations for video games.
From the perspective of computational electromagnetism, there’s interest in being able to design meta surfaces and meta materials that produce a particular response in particular applications. It’s difficult to simulate such materials, and optimizing materials for a particular optical response is a difficult problem, known as inverse rendering, in computer graphics. Work in this area is very much a core element of my current research. As an application, consider radar in self-driving cars. You want to be able to simulate what the sensors see but do so not on the road in a real physical environment, which can be dangerous. You might want to do so to train your software and validate it. You might want to optimize materials that could interfere with and blind your sensors, or design your sensors or software in a way that is more resistant to such interference.
Another application of my research is in optimizing cellular communications. Say your city has a budget to build five cellular towers and you want to optimize their placement in a way that maximizes cellular coverage. Here, you would write an optimization simulation, which is extremely difficult as it needs to be a three-dimensional simulation of electromagnetic radiation. We tend to think of cell towers as objects in two dimensions, but a cellular network operates in three dimensions. Optimizing placement in a three-dimensional environment in a way that maximizes cellular coverage is the kind of problem my research aims to address.
What attracted you to the Cheriton School of Computer Science?
Waterloo has one of the top computer science schools and its computer graphics group is among the best in the world. The students here are very strong. Ultimately, what attracted me is the strength of the computer graphics group — its faculty, colleagues who are easy to work with and with whom you want to collaborate — and the quality of students at Waterloo, both the undergrads and grads.
Tell us a bit about your research.
My main research interests are in ray and wave optics, light transport and rendering to accurately model and simulate the behaviour of light and its interaction with matter in complex environments. Wave-optical light transport has been approached from a computer graphics perspective, but usually lacking rigour and accuracy, as well as from the optical side, but often not efficiently enough to tackle real-life environments. My unique research focus addresses efficient, but accurate, wave-optical light transport. The goal is to develop a cohesive computational theory that allows the simulation of the wave-optical behaviour of light in complex, real-life scenes and environments.
Do you see opportunities for collaborative research at the School of Computer Science?
Yes, obviously within the computer graphics group there are many opportunities. Toshiya Hachisuka is one of the top light transport researchers in the world and he is a member of our computer graphics group. He is an obvious person with whom my students and I could collaborate. But there are many opportunities with other researchers at the School of Computer Science and with experts in optics from engineering and physics at Waterloo.
What do you see as your most significant contribution?
I’d say my line of research on physical light transport, which accounts for the wave nature of light globally in a scene and is able to reproduce wave-interference and diffraction effects of real physical objects, is my most significant contribution. It’s not a single paper per se, but rather a series of papers I published from 2021 to 2023 that in some sense solves the problem of wave-optical light transport. This work draws a concrete connection between the way we do light transport for computer graphics and modern wave-optical formulations. The tools I developed for rigorous wave optics are novel and quite powerful.
What do you do in your spare time?
I’m a powerlifter, a strength sport that focuses on lifting maximum weight on three different kinds of lifts — squat, bench press and deadlift. I’ve been doing powerlifting for years.
I also enjoy history, both modern history and older European history, and I enjoy reading. I’ve worked in the computer games industry, and have in interest in video games, though it’s hard to find the time to play them now. I also enjoy travelling, hiking and exploring our planet.