Seminars/Colloquia/Invited Talks

Seminars

Yaser Sheikh

Bayesian Modeling of Dynamic Scenes for Object Detection

PLACE:	Clark 314
EVENT:	CIS Seminar Series
DATE:	July 19, 2005
TIME:	1:00-2:00

Abstract

In this talk, we present a novel probabilistic model of images (or video) as a distribution in R5. By using a non-parametric (model-free) density estimation method over a joint domain-range representation of image pixels, complex dependencies between the domain (location) and range (color) are directly modeled. We demonstrate the efficacy of this model by applying it to the detection of objects in the presence of dynamic backgrounds, modeling both the foreground and the background in a coherent manner. The background and foreground models are then used competitively in a MAP-MRF decision framework, stressing spatial context, and we show that the posterior function can be maximized efficiently by finding the minimum cut of a capacitated graph. 1. Bayesian Modelling of Dynamic Scenes for Object Detection, Yaser Sheikh and Mubarak Shah, IEEE Transactions on Pattern Analysis and Machine Intelligence, (accepted). 2. Object Tracking Across Multiple Independently Moving, Yaser Sheikh and Mubarak Shah, IEEE International Conference on Computer Vision, 2005. 3. Exploring the Space of an Action for Human Action Recognition, Yaser Sheikh, Mumtaz Sheikh and Mubarak Shah, IEEE International Conference on Computer Vision, 2005. 4. Bayesian Object Detection in Dynamic Scenes, Yaser Sheikh and Mubarak Shah, IEEE Conference on Computer Vision and Pattern Recognition, 2005. (Oral)

Brief biography

Yaser Sheikh received the B.S. degree in Electronic Engineering from the Ghulam Ishaq Khan Institute of Engineering Sciences and Technology, Pakistan in 2001 and is currently a doctoral student at the University of Central Florida (UCF). He is a recipient of the Hillman Fellowship Award in 2004 for excellence in research in Computer Science. His current research interests include Bayesian Modeling for Computer Vision, Co-operative Sensing, Human Action Recognition and Visual Analysis of Aerial Imagery. http://www.cs.ucf.edu/~yaser/