Back to the program.
Multisensor Data Fusion for Simultaneous People Tracking and Recognition with a Service Robot
Nicola Bellotto
University of Essex
Dept. of Computing and Electronic Systems
nbello@essex.ac.uk

Abstract

In recent years, different solutions have been implemented on mobile robots to perform, independently, human tracking and recognition. Better performances can be achieved, instead, when the two tasks are accomplished simultaneously. In this talk, a new architecture will be presented for joint people tracking and recognition with a mobile service robot, which includes sensor data fusion and Bayesian estimation. The robot integrates the information provided by legs and face recognition algorithms to locate, identify and finally approach possible communication partners. Experimental results and the application to a receptionist robot will be discussed.

Keywords

Human tracking and recognition, Mobile service robotics, Sensor fusion