Seminars/Colloquia/Invited Talks

Seminars

Jana Kosecka

Vision Based Mapping and Localization

PLACE:	Clark 314
EVENT:	CIS Seminar Series
DATE:	November 23, 2004
TIME:	1:00 - 2:00

Abstract

Vision based localization is central to basic navigation and map building tasks. In the presence of large scale environments, in addition to environment geometry, the knowledge of environment topology can be efficiently exploited in the context of these tasks. The topological model is inferred in the exploration stage and consists of a set of locations and neighbourhood relationships between them. By representing the individual locations with local geometric features and discriminative appearance-based signatures enables us to solve the problem of global localization by means of location recognition. The misclassifications due to dynamic changes in the environment or inherent appearance ambiguities are overcome by exploiting the neighbourhood relationships captured by a Hidden Markov Model. Once the most likely location has been determined we demonstrate how to recover the relative pose of the camera with respect to the model and the local metric structure of the environment.

Brief biography

Jana Kosecka is an Assistant Professor at the Department of Computer Science, George Mason University. She obtained her M.S.E. in Electrical Engineering and Computer Science from Slovak Technical University and Ph.D. in Computer and Information Science from University of Pennsylvania, Philadelphia as a member of GRASP laboratory. In 1996 -1999 she was a Research Scholar at University of California, Berkeley, affiliated with Robotics Laboratory and PATH. She is the recipient of David Marr's prize (with Y. Ma, S. Soatto and S. Sastry) for Euclidean reconstruction and reprojection up to subgroups. She received the National Science Foundation CAREER Award and is Associated Editor of IEEE Transactions on Robotics and International Journal of Computer Vision.  Her research interest include computer vision, robotics and hybrid control and sensing for distributed systems. In particular she is interested 'seeing' systems engaged in autonomous tasks and acquisition of static and dynamic models of environments be means of visual sensing.