Evaluating Human-Robot Interaction from Inside and Outside Comparison Between the First-Person and the Third-Party Perspectives

Kazuki Yanai
Waseda University, Japan

Aiko Murata
Waseda University, Japan / Japan Society for the Promotion of Science, Japan

Ryota Mizutani
Research & Technology Group, Fuji Xerox Co., Ltd., Japan

Akira Ichiboshi
Research & Technology Group, Fuji Xerox Co., Ltd., Japan

Kazunari Komatsuzaki
Research & Technology Group, Fuji Xerox Co., Ltd., Japan

Roshan Thapliya
Research & Technology Group, Fuji Xerox Co., Ltd., Japan

Katsumi Watanabe
Waseda University, Japan / The University of Tokyo

Ladda ner artikel

Ingår i: KEER2018, Go Green with Emotion. 7th International Conference on Kansei Engineering & Emotion Research 2018, 19-22 March 2018, Kuching, Malaysia

Linköping Electronic Conference Proceedings 146:30, s. 276-286

Visa mer +

Publicerad: 2018-03-13

ISBN: 978-91-7685-314-6

ISSN: 1650-3686 (tryckt), 1650-3740 (online)


Evaluating the quality of human-robot interactions is essential for designing social robots. To design an integrated system with robots and people working together in an office or school, it is important to consider not only how a person feels when directly interacting with a robot (firstperson perspective), but also how people evaluate the interaction from a third-person perspective. In this study, we aim to identify the factors in human-robot interaction that improve communication from the first-person and the third-person perspectives and examine the relationship between these factors. First, we asked participants to interact with a robot and videotaped their interactions. After the interaction session, the participants completed questionnaires for evaluating the interaction from a first-person perspective. Next, we showed this video to another group of participants and asked them to evaluate the robot-participant interactions. This was done to get a third-person perspective. The third-person perspective evaluations were mostly consistent among the evaluating participants. On the other hand, the third-person evaluations did not necessarily match those from the first-person perspective. However, several factors in the first-person evaluations correlated with how people would have a good impression toward the observed interaction. The results suggest that certain factors contribute to forming consistent impressions of human-robot interactions from the thirdperson perspective.


Human-robot interaction, Interaction evaluation, Third-person perspective


Inga referenser tillgängliga

Citeringar i Crossref