March 30, 2021

12:00 pm / 1:00 pm

Venue

ZOOM

Recorded Seminar:

https://wse.zoom.us/rec/share/PEq3loKUV7T9INRUsypaFX1D1SpTXIScM7m2WG3jypGVoYyYvUUBuWRn5sthaVoU.GU0OrYdzqvkf6dEL?startTime=1617119237000

Juan Carlos Niebles, PhD
Stanford AI Lab

“EventUnderstanding: a Cornerstone of Visual Intelligence”

 Abstract:  Asevents are central to human life, computer vision systems of the future willneed to recognize and understand human events at various levels of detail. Manyof these events are naturally organized by part-of hierarchical relationshipsin a partonomy. We argue that leveraging such partonomy structure will be keyto the success of human event understanding algorithms. In this talk, wesummarize our recent progress in algorithms for recognizing and parsing eventsthat go from simple atomic actions to complex long term tasks. At the scaleofatomic actions, we will discuss an efficient algorithm for video actionclassification that leverages spatio-temporal shift operations instead of convolutions.In the intermediate scale, we study actions as compositions ofspatio-temporalscene graphs and show how they can both improve action recognition and enablebetter few-shot learning of actions. When such supervisedcompositions are notavailable, we show how to leverage data-driven spatio-temporal graphs for videocaptioning. At the long term task scale, we discuss a method for learning the structure of tasks from instructional videos and show its applications toprocedure planning. We conclude by discussing some of the open challenges thatremain to fully embrace the partonomy perspective for event understanding

 Bio:  Juan Carlos Niebles received an Engineering degree in Electronics from Universidaddel Norte (Colombia) in 2002, an M.Sc. degree in Electrical and Computer Engineering from University of Illinois at Urbana-Champaign in 2007, and a Ph.D. degree in Electrical Engineering from Princeton University in 2011. He is co-Director of the Stanford Vision and Learning Lab, Associate Director of Research at the Stanford-Toyota Center for AI Research and a Senior Research Scientist at the Stanford AI Lab since 2015. He was also an Associate Professor ofElectrical and Electronic Engineering in Universidad del Norte (Colombia)between 2011 and 2019. His research interests are in computer visionandmachine learning, with a focus on visual recognition and understandingof humanactions and activities, objects, scenes, and events. He has served as AreaChair for the top computer vision conferences CVPR and ICCV. He is also amember of the AI Index Steering Committee and is the Curriculum Director forStanford-AI4ALL. He is a recipient of a Google Faculty Research award (2015),the Microsoft Research Faculty Fellowship (2012), a Google Research award (2011)and a Fulbright Fellowship (2005).