FlowGames

Abstract

Computer vision-based interfaces to games hold the promise of rich natural interaction and thus a more realistic gaming experience. Therefore, the video games industry started to develop and market computer vision-based games recently with great success. Due to limited computational resources, they employ mostly simple algorithms such as background subtraction, instead of sophisticated motion estimation or gesture recognition methods. This not only results in a lack of robustness, but also in very limited interaction possibilities and thus reduced gaming experience. In this paper, we show a couple of concepts to control video games based on optical flow. We use a state-of-the-art optical flow algorithm able to be computed densely in real-time on GPUs, which are in fact built-in in nearly every gaming hardware available. Based on the estimated motion, we develop several computer games with increasing complexity: Starting with using the flow field as force acting on moveable objects, we span the spectrum to more sophisticated concepts such as controlling widgets and action recognition.