Talk: Learning Actions from Humans in Video
Modeling & understanding actions is key for computer vision
Monday, November 27, 2023 · 4 - 5:15 PM
Hybrid
·
Engineering : 231
Advances in Perception, Prediction, and Reasoning
Learning Actions from Humans in Video
4:00-5:15pm ET, Monday, Nov 27, 2023
UMBC, Engineering 231 and via WebEx
Eadom Dessalene
University of Maryland, College Park
The prevalent computer vision paradigm in the realm of action understanding is to directly transfer advances in object recognition toward action understanding. In this presentation, I discuss the motivations for an alternative embodied approach centered around the modeling of actions rather than objects and survey recent work of ours along these lines, as well as promising possible future directions.
Eadom Dessalene is a Ph.D. Candidate at the University of Maryland, College Park, advised by Yiannis Aloimonos and Cornelia Fermuller in the Perception and Robotics Group. Eadom received his bachelor's degree in Computer Science from George Mason University. He has made several important contributions to research on video understanding, ego-centric vision, and action understanding through publications in CVPR, ICLR, T-PAMI, and ICRA, as well as winning first place in the 2020 EPIC Kitchens Action Anticipation Challenge.
The Advances in Perception, Prediction, and Reasoning (PPR) talks are organized and hosted by UMBC Professor Tejas Gokhale.