How Can Robot's Gaze Ratio and Body Direction Show an Awareness of Priority to the People With Whom it is Interacting?
How Can Robot's Gaze Ratio and Body Direction Show an Awareness of Priority to the People With Whom it is Interacting?
复制标题
DOI：
10.1109/lra.2019.2929992
复制
发表时间：
2019-10
影响因子：
5.2
通讯作者：
Honoka Arai;Mitsuhiko Kimoto;T. Iio;K. Shimohara;Reo Matsumura;M. Shiomi
中科院分区：
文献类型：
--
作者：
Honoka Arai;Mitsuhiko Kimoto;T. Iio;K. Shimohara;Reo Matsumura;M. Shiomi
当一个人与需要不同权限级别的其他人交互时，每个人的他/她的注视比率和身体方向是不同的，例如，一个销售人员转向一个非常重要的人（VIP），比VIP的工作人员更专注。在这项研究中，我们调查的凝视比例和身体方向的影响，社会机器人表现出权威意识时，他们同时与多个人互动。我们为社交机器人开发了一个凝视控制器系统，并通过实验研究了当机器人销售人员向扮演不同角色的两个人描述物品时凝视比率和身体方向（优先级导向：身体方向转向目标和非优先级导向：身体方向不转向目标）的影响：VIP和追随者（例如，VIP的工作人员）。我们的实验结果表明，不同的优势的注视比例和身体方向表现出的优先意识。在优先级导向的条件下，参与者更高度评价VIP和追随者之间100：0的凝视比例。但是，在非优先导向的条件下，参与者对80：20或90：10的注视比例的评价更高。这些结果有助于社交机器人的凝视行为设计，与多个人谁需要不同的优先级。
When a person interacts with others who require different levels of authority, his/her gaze ratio and body direction for each person are different, e.g., a salesperson who turns to and looks at a very important person (VIP) more attentively than a member of that VIP's staff. In this study, we investigate the effects of gaze ratios and body directions with which social robots demonstrate an awareness of authority when they are simultaneously interacting with multiple persons. We develop a gaze-controller system for a social robot and experimentally investigate the effects of the gaze ratio and the body direction (priority oriented: body direction turns to a target and non-priority oriented: body direction does not turn to a target) when the robot salesperson is describing items to two persons who are playing different roles: a VIP and a follower (e.g., member of a VIP's staff). Our experiment results show that different advantages of gaze ratio and body direction demonstrate an awareness of priority. In the priority-oriented condition, participants more highly evaluated the gaze at 100:0 ratio between VIPs and followers. But, in the non-priority-oriented condition, the participants more highly evaluated the gaze ratios at 80:20 or 90:10. These results contribute to a gaze behavior design for social robots that interact with multiple persons who require different levels of priority.