Combining Sound Localization and Laser-based Object Recognition

Abstract

Mobile robots, in general, and service robots in human en-vironments, in particular, need to have versatile abilities to perceive and interact with their environment. Biologically in-spired sound source localization is an interesting ability for such a robot. When combined with other sensory input both the sound localization and the general interaction abilities can be improved. In particular, spatial filtering can be used to im-prove the signal-to-noise ratio of speech signals emanating from a given direction in order to enhance speech recognition abilities. In this paper we investigate and discuss the com-bination of sound source localization and laser-based object recognition on a mobile robot.