Body Movement Synchrony Captured by an Omnidirectional Camera predicts the Degree of Information Transfer during Dialogue: Toward Automatic Evaluation of Verbal Communication

摘要

In a meeting, two or more individuals come together to verbally share information. However, discrepancies in perceptions often occur in complicated discussions. Therefore, whether the information transfer succeeded must be reconfirmed. This problem is typically resolved through ex-post measures such as meeting minutes. Our previous results indicated that the degree of information transfer can be predicted using head movement synchrony. In this study, we propose an information transfer estimation method based on head movement synchrony during a conversation evaluated by an omnidirectional camera that can easily obtain body movements. These body movements were quantified using Motion Energy Analysis, which is commonly used in clinical psychology, with additional processes used to reduce noise caused by lighting in the room and resulting from large accidental movements. By these processes, body movement synchrony was quantified. We confirmed that this synchrony occurred between actually interacting pairs and tested whether it was related to the degree of information transfer. In the experimental evaluation, body movement synchrony was found to be high for interacting pairs, and a positive correlation was observed between synchrony and degree of information transfer. Therefore, our proposed method may detect the synchrony and estimate the degree of information transfer using an omnidirectional camera.