One-to-One Audio Guidance System Using Human Vision, Designed for a Guide Robot

In this study, we present a guide robot that speaks to humans through a directional loudspeaker. The robot is equipped with an image-sensing device (human vision component) that acquires the age, gender, and direction of the face. The robot rotates in the direction of the human face and speaks, the contents of the speech depend on the information received from the human vision component. The directional loudspeaker is a parametric speaker installed with an ultrasonic speaker array. The system was constructed using the RT middleware. The robot can speak to a single person within a narrow audible area. This paper discusses the characteristics of the developed system.