A model of horizontal 360° object localization based on binaural hearing and monocular vision

Abstract

We introduce a biologically inspired localization system, basedon a “two-microphone andone camera” configuration. Our aim is to perform a robust, multimodal 360◦ detection of objects, in particular humans, in the horizontal plane. In our approach, we consider neurophysiological findings to discuss the biological plausibility of the coding and extraction of spatial features, but also meet the demands and constraints of a practical application in the fieldof human-robot interaction. Presently we are able to demonstrate a model of binaural sound localization using Interaural Time Differences for the left/right detection andsp ectrum-basedfeatures to discriminate between in front and behind. The objective of the current work is the multimodal integration of different types of vision systems. In this paper, we summmarize the experiences with the design and use of the auditory model and suggest a new model concept for the audio-visual integration of spatial localization hypotheses.