Starting from the assumptions that human-ambient intelligence interaction will be improved by having more human-human like communications and that facial expressions are fundamental in human-human communications, we describe how we developed facial expressions for artificial agents based on a psychological theory developed by Scherer. In this current article we describe briefly the psychological theory that we have chosen as well as some of the reason for adopting it. We then describe the two different platforms we used, the Cherry avatar and the iCat Robot with a particular focus on the robot. Finally we explore the steps needed to adapt the psychological theory to the two different platforms and we present some conclusions and future development.
Marco Paleari, Amandine Grizard, Christine L. Lise