Abstract: Through research in Computer Vision and Deep Learning over the last few decades we have developed algorithms for successfully detecting objects and inferring their structure from images. Many such algorithms use multiple views of the same scene to infer a three-dimensional model of the world from two-dimensional images. I will review some of these capabilities in the large, reconstructing large-scale scenes to support applications such as virtual reality and autonomous vehicles, and in the small, to estimate the structure of biological molecules such as viruses and proteins, at atomic resolutions.
David Fleet: Professor, Department of Computer and Mathematical Science, University of Toronto Scarborough, CIFAR AI Chair, Faculty Member, Vector Institute, Senior Research Scientist at Google Brain Research. Prior to joining the University of Toronto in 2003 he managed the Digital Video Analysis, and Perceptual Document Analysis, Groups at the Palo Alto Research Center (PARC), and was on faculty at Queen's University. His research spans visual neuroscience, computer vision, image processing and machine learning, for which he has won numerous awards, including the Alfred P. Sloan Research Fellowship, the Koenderink Prize, and an honourable mention for the David Marr Prize. He has served as Associate Editor (2000-2003) and Associate Editor-in-Chief (2004-2008) for IEEE Transactions on Pattern Analysis and Machine Intelligence (PAMI). He has also served as Program Co-Chair of IEEE CVPR (2003), ECCV (2014). He was Senior Fellow of the Canadian Institute for Advanced Research (2005-2019), snf Associate Research Director, Industry Innovation of the Vector Institute (2019-2020).