EMOTION: Expressive Motion Sequence Generation for Humanoid Robots With In-Context Learning

摘要

This paper introduces a framework, called <monospace xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">EMOTION</monospace>, for generating expressive motion sequences in humanoid robots, enhancing their ability to engage in human-like non-verbal communication. Non-verbal cues such as facial expressions, gestures, and body movements play a crucial role in effective interpersonal interactions. Despite the advancements in robotic behaviors, existing methods often fall short in mimicking the diversity and subtlety of human non-verbal communication. To address this gap, our approach leverages the in-context learning capability of large language models (LLMs) to dynamically generate socially appropriate gesture motion sequences for human-robot interaction. We use this framework to generate 10 different expressive gestures and conduct online user studies comparing the naturalness and understandability of the motions generated by <monospace xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">EMOTION</monospace> and its human-feedback version, <monospace xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">EMOTION++</monospace>, against those by human operators. The results demonstrate that our approach either matches or surpasses human performance in generating understandable and natural robot motions under certain scenarios. We also provide design implications for future research to consider a set of variables when generating expressive robotic gestures.