Modeling situation-dependent nonverbal expressions for a pair of embodied agent in a dialogue based on conversations in TV programs

  • Authors:

  • Keita Okuuchi;Koh Kakusho;Takatsugu Kojima;Daisuke Katagami

  • Affiliations:

  • Kwansei Gakuin University, Sanda, Japan;Kwansei Gakuin University, Sanda, Japan;Shiga University of Medical Science, Otsu, Japan;Tokyo Polytechnic University, Tokyo, Japan

  • Venue:
  • HCI'13 Proceedings of the 15th international conference on Human-Computer Interaction: interaction modalities and techniques - Volume Part IV
  • Year:
  • 2013

Quantified Score

Hi-index 0.00

Visualization

Abstract

Mathematical model for controlling nonverbal expressions of a pair of embodied agents designed for presenting various information through their dialogue is discussed. Nonverbal expressions of a human during conversation with others depend on those of them as well as the situation of the conversation. The proposed model represents the relationship between nonverbal expressions of a pair of embodied agents in different situations of conversation by a constraint function, so that the nonverbal expression of each agent reproduces the characteristic of nonverbal expressions observed in human conversation with various situations in TV programs by minimizing the function.