TITLE: Tracking People and Learning Typical Motion Behaviours of Persons with Mobile Robots

PRESENTER: Professor Wolfram Burgard    http://www.informatik.uni-freiburg.de/~burgard/

AFFILIATION: Institut fur Informatik, Albert-Ludwigs-Universitat Freiburg

DATE: Friday 26 April 2002

TIME: 12:00 noon - 1:00pm

PLACE: Seminar Room K17

ABSTRACT:

One of the goals in the field of mobile robotics is the development of mobile platforms which operate in populated environments and offer various services to humans. For many tasks it is highly desirable that a robot can determine the positions of the humans in its surrounding. In this talk we present a sample-based variant of Joint Probabilistic Data Association Filters and its application to tracking the motions of people in the vicinity of the robot. We present several experiments illustrating the robustness of our approaches even in dynamic environments. Additionally we describe how the information provided by our people tracker can be exploited to learn accurate 2d and 3d maps in populated environments. Finally, we present a technique that allows a robot to learn typical motion behaviours of persons. Data, recorded by mobile robots equipped with laser range finders, is clustered into different types of motion using the popular expectation maximisation algorithm, while simultaneously learning multiple motion patterns. Experimental results, obtained using data collected in a domestic residence and in an office building, illustrate that highly predictive models of human motion patterns can be learned.

BIOGRAPHY OF SPEAKER:

Wolfram Burgard studied Computer Science at the University of Dortmund, Germany. After receiving his Ph.D. from the University of Bonn in 1991, he was a research assistant at the University of Bonn. Since 1999 he is ass ociate professor at the University of Freiburg, Germany, where he is heading the research lab for Autonomous Intelligent Systems. His research mainly focuses on the development of robust and adaptive techniques for robot state estimation and control. Over the past years his group and he have developed a series of innovative probabilistic techniques for robot navigation and control. They cover different aspects such as localisation, map-building, path-planning, and exploration. In 1997 his group deployed Rhino as the first interactive mobile tour-guide robot in the Deutsches Museum Bonn in Germany. During its deployment period of six days Rhino guided thousands of visitors through the crowded museum.

Seminar Convenor:

Bernhard Hengst
Tel: +61-2-9385-3988,
E-mail: bernhardh@cse.unsw.edu.au
School of Computer Science & Engineering, UNSW.