Intelligent three-dimensional vision sensor with ears

This paper describes our newly developed intelligent robot sensor system which comprises two eyes and four ears on a movable head. It can acquire a world image, i.e., a dynamical 3-D map of the environment, while showing humanlike behavior naturally and autonomously. The most important feature of the sensor system is the original and unified sensory architecture of the low-level and intermediate-level visual and auditory functions. We describe in this paper the key items for realizing this sensor: 1) features of the unified sensory architecture ''gradient correlation sensing principle,'' 2) the design of sensory probes and circuits for the vision section, 3) the design of sensory probes and circuits for the auditory section, 4) various algorithms to extract sensory information from each sensor, and 5) autonomous and active sensing procedures by means of visual-auditory fusion and sensory-motor control.