         
|
GRASP Seminar Series: Spring 2005
April 29, 11:00 AM, Levine Hall 307.
Ali Shokoufandeh
Computer Science Dpt, Drexel University
“Many-to-Many Feature Matching in Object Recognition”
Abstract: One of the bottlenecks of current recognition systems is their assumption of one-to-one features correspondence. This assumption breaks down in the generic object recognition task where, for example, a collection of features at one scale (in one image) may correspond to a single feature at a coarse scale (in the second image). Generic object recognition therefor requires the ability to match features many-to-many. This talk will present an approach for many-to-many matching of scale-space features. We will also present an approximation method for selecting canonical features that can be effectively used to increase the robustness of the many-to-many feature matching.
Biography: Ali Shokoufandeh received his BSc degree in computer science from University of Tehran in 1990. He joined Rutgers University in 1993 for his graduate studies in computer science (MSC in 1996 and PhD in 1999) and cognitive science (Certificate 1999). He is currently an assistant professor of computer science at Drexel University. His research focuses on applications of optimization methods in computer vision and pattern recognition.
