Search

This pilot study contributes to the building of an art installation that aims to build an emotional communication loop with the audience. It will do so by reacting to or mimicking the audience's changes in emotional expressions according to emotional contagion dynamics. The study aims to inform the project by gaining a better understanding of emotional contagion patterns and the factors that may affect how people emotionally engage in this art context. The analysis of our early experiments shows reflex mechanisms of facial expression mimicry and counter-mimicry that follow patterns similar to those reported in the psychology literature. In fact, automatic mimicry and counter-mimicry correlated to some extent to whether or not the audience felt to be interacting with a real person. Furthermore, the results indicate that individual differences play a role in the way people can emotionally engage with this type of artwork. However, irrespective of these differences, the interaction led the audience to introspect and reflect about emotions.

To endow virtual agents with more realistic affective behavior, the notion of expectation-based emotions plays an important role: emotional states of agents should not only be triggered by present stimuli, but also by anticipation on future stimuli, and evaluation of past stimuli in the context of these anticipations. Within this study, an extension of the belief-desire-intention (BDI) model with expectation-based emotions is proposed. The model has been implemented in the modeling language LEADSTO. In addition, a game application has been developed, in which a user can play a dice game against an agent that is equipped with the emotion-based model. An empirical evaluation indicates that the model significantly enhances the agent's believability, in particular concerning its involvement in the situation.