An Affective Agent Framework Towards Believable Agents for Digital Entertainment

Digital entertainment industry uses multi-agents techniques based on several different approaches to establish autonomous characters that behave the most believable way as possible. This article presents a preliminary version of an agent architecture based on psychological and neurobiological theories. We expect that this architecture could facilitate implementation of synthetic believable characters by providing ways of defining different personalities for them while keeping the decisionmaking process evaluation time so short that allows its use in real-time applications.