Understanding Dynamic Social Grouping Behaviors of Pedestrians

There have been many studies in the literature on social group recognition of crowds of pedestrians. However, most of these studies have approached the problem from a static point of view. A study on the dynamic property of social groups among people over time can provide significant insight into human behaviors and events. Inspired by sociological models of human collective behavior, in this work, we present a framework for characterizing hierarchical social groups based on evolving tracklet interaction network (ETIN) where the tracklets of pedestrians are represented as nodes and the their grouping behaviors are captured by the edges with associated weights. We use non-overlapping snapshots of the interaction network and develop the framework for a unified dynamic group identification and tracklet association. The approach is evaluated quantitatively and qualitatively on videos of pedestrian scenes where manually labeled ground-truth is given. The results of our approach are consistent to human-perceived dynamic social groups of the crowd. The performance analysis of our method shows that the approach is scalable and it provides situational awareness in a real-world scenarios.