Kyros Kutulakos, a professor in the Department of Computer Science, is a recipient of a 2023 Dean’s Research Excellence Award from the Faculty of Arts & Science.
Professor Kyros Kutulakos is a recipient of a 2023 Dean’s Research Excellence Award from U of T’s Faculty of Arts & Science.
The Arts & Science Dean’s Research Excellence Award recognizes mid-career researchers with “significant and sustained research achievement and impact.”
Kutulakos has produced original contributions to the fields of computer vision and computational photography and has received numerous best paper awards at premier conferences, such as the International Conference on Computer Vision (ICCV), the European Conference on Computer Vision (ECCV), and Computer Vision and Pattern Recognition (CVPR).
Kutulakos’ research comprises deep theoretical discoveries and disruptive new visual sensing techniques and is widely admired for its creativity, rigour and transformational impact.
This includes highly influential work concerning the estimation of 3D structures from multiple 2D photos of a scene. While it was well known that multiple views of a scene provide information about 3D geometry and appearance of objects, fundamental questions had no answers. In what sense do n photos of a scene determine 3D shape, and what are the inherent ambiguities? Do provably correct reconstruction algorithms exist for general scenes? Kutulakos and Steve Seitz introduced a mathematical framework to answer these and other long-standing questions, along with a state of the art-algorithm called Space Carving. This work received the 1999 David Marr Prize, a research prize awarded to the best paper at ICCV and laid the foundation for modern 3D scene reconstruction methods.
Other seminal and award-winning contributions throughout his career include publishing a theory of computational light transport that uses cameras to capture many of the complex ways in which light interacts with the environment; proposing the first effective method for low-power 3D visual sensing in bright sunlight and other challenging lighting conditions; and showing that with everyday cameras, one can measure properties of the electrical grid.
His recent research includes developing what he calls the “ultimate video camera”: a camera whose pixels can record and analyze the precise arrival time of every incoming photon. Alongside his PhD student Mian Wei, postdoctoral fellow Sotirios Nousias and Assistant Professor David Lindell, they show that such cameras make it possible to image dynamic scenes over an extreme range of timescales. In practice, this has led to the very first experimental demonstration of ultra-wideband video recording: video of a dynamic scene that can be played back at 30 Hz to show everyday motions but can also be played a billion times slower to show the propagation of light itself. These results have the potential to spark significant advances on multiple fronts in computer vision, 3D sensing and scientific imaging. This research was awarded the David Marr Prize at ICCV 2023.
“I am both thrilled and humbled to receive this award,” said Kutulakos. “My wonderful students, postdocs and colleagues over the years deserve a great deal of credit for this. It makes me proud to be part of an institution that values long-term, curiosity-driven research.”
Eyal de Lara, professor and chair of the Department of Computer Science, praised Kutulakos for his sustained research excellence.
“Kyros has made profound contributions to computer vision and computational photography and is highly regarded among his peers for his continued and consistently cutting-edge work,” said de Lara. “His exceptional achievements, along with his remarkable record of service to the research community, make him highly deserving of this award.”