These days, established computer vision professors are given to complaining, with varying degrees of seriousness, that current Ph.D. students do not know any work in the field that pre-dates the “deep learning revolution” of 2012. However, while wholesale amnesia is unquestionably dangerous for the field, from a pragmatic point of view, even the “old guard” concedes that it is no longer necessary to teach historic work that was truly an intellectual dead end. This short course is an attempt to grapple with the question of what “classical” computer vision techniques should be considered a “must know” for researchers entering the field today, and how past trends and approaches should inform the field as it looks poised to enter a challenging phase—continuing its spurt of rapid growth even while the initial momentum from the “deep learning revolution” begins to fade and negative societal impacts of some maturing technologies come into view.
Week 1 (January 28-30)
Introduction; Historical Overview—Lecture Slides
Tuesday, January 28, 2020
3:00 pm–4:30 pm
TSRB Auditorium 118
Connections to Cognitive Science and Psychophysics—Lecture Slides
Wednesday, January 29, 2020
3:00 pm–4:30 pm
TSRB Auditorium 118
History of Ideas in Recognition: Part I—Lecture Slides
Thursday, January 30, 2020
3:00 pm–4:30 pm
TSRB Auditorium 118
Week 2 (February 4–6)
History of Ideas in Recognition: Part II—Lecture Slides
Tuesday, February 4, 2020
3:00 pm–4:30 pm
TSRB Banquet Hall
Future Trends
Wednesday, February 5, 2020
3:00 pm–4:30 pm
TSRB Auditorium 118
Ethical and Societal Impacts of Computer Vision Technologies
Thursday, February 6, 2020
3:00 pm–4:30 pm
TSRB Auditorium 118
Svetlana “Lana” Lazebnik is an associate professor in the Department of Computer Science at the University of Illinois at Urbana-Champaign (U of I). She received her Doctor of Philosophy degree in Computer Science at U of I in 2006 and was an assistant professor at the University of North Carolina at Chapel Hill from 2007 to 2011.
Lazebnik’s research specialty is computer vision and her main research themes include scene understanding, joint modeling of images and text, and deep learning techniques for visual recognition problems. In recent years, her research group has worked on fundamental “building blocks” for image-language understanding, including two-branch neural networks for matching images and sentences; and datasets and models for visual grounding, or localizing visual concepts described by phrases in image regions.
Lazebnik’s notable awards include the 2016 Longuet-Higgins Prize at CVPR 2016 for a CVPR 2006 paper with significant impact on computer vision; the 2013 Sloan Research Fellowship; the 2009 Microsoft Research Faculty Fellowship; and the 2008 NSF CAREER Award. Lazebnik is currently editor in chief of the International Journal of Computer Vision. She served as program chair for the 2012 European Conference on Computer Vision and the 2019 International Conference on Computer Vision.
]]>