Host: The Japan Society of Mechanical Engineers
Name : [in Japanese]
Date : September 19, 2023 - September 21, 2023
In this research, we developed and evaluated a listening system that guides the line of sight to images presented by speech using a familiar robot as a dialogue agent. The presentation image has a function of nodding in accordance with the utterance rhythm. In this system, we conducted experiments using two modes. One is that subject communicate with the listener’s character on the display and guide the listener’s character to gaze at the image. The other that subject communicate face-to-face with the robot and guides the robot to gaze at the image. In both modes, the presented image either nods in accordance with the utterance rhythm or does not nod at all. The experiment was conducted on 31 elderly people. As a result of two-factor analysis of variance, significant differences were observed between the listener’s character and the robot in the items of enjoyment, familiar and ease of speaking and the robot was evaluated highly.
