Adapting virtual camera behaviour through player modelling

Abstract

Research in virtual camera control has focused primarily on finding methods to allow designers to place cameras effectively and efficiently in dynamic and unpredictable environments, and to generate complex and dynamic plans for cinematography in virtual environments. In this article, we propose a novel approach to virtual camera control, which builds upon camera control and player modelling to provide the user with an adaptive point-of-view. To achieve this goal, we propose a methodology to model the player’s preferences on virtual camera movements and we employ the resulting models to tailor the viewpoint movements to the player type and her game-play style. Ultimately, the methodology is applied to a 3D platform game and is evaluated through a controlled experiment; the results suggest that the resulting adaptive cinematographic experience is favoured by some player types and it can generate a positive impact on the game performance.