A high-level architecture for believable social agents

Abstract

Abstract:The creation of virtual humans capable of behaving and interacting realistically with each other requires the development of autonomous believable social agents. Standard goal-oriented approaches are not well suited to it because they don't take into account important characteristics identified by the social sciences. This paper tackles the issue of a general social reasoning mechanism, discussing its basic functional requirements using a sociological perspective, and proposing a high-level architecture based on Roles, Norms, Values and Types.

The Emergence of Autonomous Representations in Artificial Agents

“…The majority of artificial agents inhabiting into artificial environments are designed using architectures that are based on the cognitivist assumption [5], [6]. However, there are a lot of practical problems.…”

Lets Come Together – Social Navigation Behaviors of Virtual and Real Humans

“…IPA is based on a classification of social interactions that take place in small groups and essentially distinguishes between socio-emotional factors that refer to the social relationships within a group, such as positive or negative feedback to group members, and task-oriented factors that refer to group tasks, such as asking questions or summarizing and offering direction. It has already been successfully employed in other system of social group dynamics, see for example [5] or [15].…”

“…In contrast, Thalmann and colleagues [5] concentrate on the simulation of social navigation behaviours in virtual 3D environments including the social avoidance of collisions, intelligent approach behaviours, and the calculation of suitable interaction distances and angles. The work is based on an operationalization of empirically-grounded theories of human group dynamics, such as Kendon's group formation system [9].…”

“…Unlike Thalmann and colleagues [5] who focus on the simulation of human navigation behaviors using virtual agents, we involve the users as active participants in the scenario. In particular, we allow them to freely navigate through the scenario and join or leave groups of other agents by making use of a commercially available dancing pad that is employed in many computer games.…”

Theatre-outside-of-theatres : spaces of digital performance

“…Several agent architectures for Interactive Drama and video games have, in fact, adopted this model for making believable characters (cf Damiano et al 2005;. Sandercock et al 2006;Padgham & Winikoff 2004;Bratman et al 1991;Guye-Vuilleme & Thalmann 2000). Similarly, Ryan's possible worlds model, where characters make multiple conjectures about the future, bears strong similarities to recent speculations in cognitive science that the cortex is a kind of Bayesian prediction machine.…”