Over the past years, extensive research has been dedicated to developing robustplatforms and data-driven dialog models to support long-term human-robot interactions.However, little is known about how people’s perception of robots and engagement withthem develop over time and how these can be accurately assessed through implicitand continuous measurement techniques. In this paper, we explore this by involvingparticipants in three interaction sessions with multiple days of zero exposure in between.Each session consists of a joint task with a robot as well as two short social chats withit before and after the task. We measure participants’ gaze patterns with a wearableeye-tracker and gauge their perception of the robot and engagement with it and the jointtask using questionnaires. Results disclose that aversion of gaze in a social chat is anindicator of a robot’s uncanniness and that the more people gaze at the robot in a jointtask, the worse they perform. In contrast with most HRI literature, our results show thatgaze toward an object of shared attention, rather than gaze toward a robotic partner, isthe most meaningful predictor of engagement in a joint task. Furthermore, the analysesof gaze patterns in repeated interactions disclose that people’s mutual gaze in a socialchat develops congruently with their perceptions of the robot over time. These are keyfindings for the HRI community as they entail that gaze behavior can be used as an implicitmeasure of people’s perception of robots in a social chat and of their engagement andtask performance in a joint task
I Can See it in Your Eyes : Gaze as an Implicit Cue of Uncanniness and Task Performance in Repeated Interactions with Robots
2021-01-01
doi:10.3389/frobt.2021.645956
Aufsatz (Zeitschrift)
Elektronische Ressource
Englisch
DDC: | 629 |
Detection and tracking of eyes for gaze-camera control
British Library Online Contents | 2004
|