Compressive Sensing, Sparse Representations and Dictionaries for Image and Video-based Recognition

Wednesday, March 21, 2012
7:00 PM
Free and open to the public

Recent developments in compressive sensing, sparse representations and dictionary learning are enabling potentially new approaches to many image and video-based recognition problems. In this talk, I will briefly review these emerging principles and discuss image and video-based face and iris recognition using sparse representations, dictionary-based object and event recognition and compressive classification of videos. Methods for handling illumination variations and low resolution images will also be discussed.

x x


Rama Chellappa

Minta Martin Professor of Engineering
University of Maryland, College Park