Studies on Grounding with Gaze and Pointing Gestures in Human-Robot-Interaction

摘要

In this study we investigated the use of gaze and pointing gestures in scenarios where a human has to follow the instructions of a humanoid robot. Our objective was to analyze the performance of a human participant, that solves an abstract jigsaw puzzle with the help of our robot instructor, in different grounding scenarios with varying difficulty. Furthermore we investigated how the attitude towards the robot and the self-assessment of the participant changed. Our results support that adding gaze to the interaction usually improves the interaction, but often additional pointing gestures are needed to make a significant difference.