Observing Attention

Understanding whether an event attracted the audience’s attention and which moments were mostly enjoyable is a primary goal for sport and show business managers. OZ (Osservare l’attenZione – Observing attention) is an interdisciplinary, mixed-methods project that aims at developing a technology able to automatically detect at run time spectators’ attention level, via the integration of microsociological analysis of human behavior into computer vision modeling and techniques. More specifically, we will show how it is possible to distinguish supporters of different teams by automatically detecting their liveliness in different moments of the match, even when they are mingled in the crowd. Moreover, we will show how, only by automatically detecting crowd’s motion on the stands, it is possible to detect and annotate the most salient events of the match, like goals, fouls or shots on goal.