Lux - An interactive receptionist robot for university open days

Abstract

Interactive service robots are ready to serve us both in public and private places soon. This paper presents a multisensor framework to the simultaneous tracking and recognition of humans so that a service robot can implement reception tasks interactively. Both a laser range scanner and a camera have been deployed in the framework. By fusing range and visual data, the robot can detect and identify applicants in the university environment in order to provide dedicated services. The robustness of the robot perception is increased by the joint tracking and recognition of face, clothes and height, implemented using a bank of filters and probabilistic data association. Experiments in the university open days demonstrate the effectiveness of the proposed solution.