string(10) "newsevents"
August 30, 2012

Intelligent Virtual Environments and Intelligent Agents: Knowledge Representation, Perception, Action, Behaviour and Development Technologies

  • May 19, 2024 till May 19, 2024
  • Main Lecture Room

Intelligent virtual environments are a relatively new scientific field that is concerned with the representation of worlds by means of computers with a focus on intelligent virtual agent behaviour. The design and development of intelligent virtual environment systems requires the convergence of a multitude of theories and technologies, in which knowledge representation, as related to either world modelling or the application of models of perception, action and behaviour generation, plays a crucial role. The subject of this presentation is the REVE representation for intelligent virtual environments. The REVE representation is a novel approach towards addressing the inherent difficulties that arise from the diversity of needs in the field, on a concrete theoretical basis, with specific focus on intelligent virtual agents and aiming at flexibility, extendibility and applicability on a variety of application domains. In addition, we present the jREVE platform, a complete, documented and fully-functional implementation of the representation accompanied by several software development components and tools. A key example of the platform’s potential is the REVE Worlds system, an integrated virtual environment with advanced virtual world visualization capabilities, support for multiple virtual agents, support for distributed deployment and operation, extendibility and the capacity to serve as a basis for further development. Moreover, we present the VERL language for the definition of virtual worlds for systems compliant with the REVE representation, as well as the SARA software development framework for intelligent virtual agents. We also briefly present the mVital system, a distributed multiagent system for logic-based intelligent agents that is capable of supporting simple virtual worlds, allows definition of agent personalities in the VAL language and supports planning-based behaviour-generation. Our presentation concludes with a discussion of use cases and applications in research and education.

Talk slides in pdf [~3,7MB]

Skip to content