Training Social Skills in Virtual Reality. Machine Learning as a Process of Co-Creation

  • Klaus Neundlinger (Autor, Hauptvortrag)
  • Michael Mühlegger (Autor, Hauptvortrag)
  • Kriglstein, S. (Autor, Hauptvortrag)
  • Thomas Layer-Wagner (Autor, Hauptvortrag)
  • Regal, G. (Autor, Hauptvortrag)

    Aktivität: Vortrag ohne Tagungsband / VorlesungPräsentation auf einer wissenschaftlichen Konferenz / Workshop

    Beschreibung

    The training of social skills in interactive Virtual Reality scenes has become an issue of rising interest in research and development. This is also the case of the project Virtual Skills Lab that deals with the question how the potential of such a training can be applied to the context of organizations. In a virtual office scene, the interaction between the player and the computer-generated agent is based on speech recognition and conversational Artificial Intelligence (AI). While studies and experiments have shown that humans are able to treat agents in Virtual Reality as if they were real persons, it is still impossible to establish truly spontaneous forms of interaction by combining speech recognition, conversational AI and Virtual Reality. Hence, existing applications for social skills training in Virtual Reality focus on the possibility to literally take over the other´s perspective or the ability to overcome social anxieties. Yet, social interaction in the workplace is always embedded in a given organizational culture and therefore characterized by behavioral patterns. In this sense, a Virtual Reality training tool for social interaction in the workplace can be enriched by design elements that enable machine learning in terms of pattern recognition. The paper outlines how the project team has dealt with the existing limitations of conversational AI in the process of working out the storyline and what design elements can be introduced in the Virtual Reality interaction in order to enable the machinic recognition of patterns of organizational culture.
    Zeitraum19 Juni 2020
    EreignistitelMedia, Arts & Design | AI 2020 Conference
    VeranstaltungstypSonstiges
    BekanntheitsgradInternational

    Research Field

    • Former Research Field - Experience Business Transformation