Video Based Children's Social Behavior Classification in Peer-Play Scenarios

Lu Tian, Dingrui Duan, Jinshi Cui, Li Wang, Hongbin Zha, Hamid Aghajan; Proceedings of the IEEE International Conference on Computer Vision (ICCV) Workshops, 2013, pp. 746-747

In the studies of developmental psychology, labeling children’s play behavior is an important process. Children’s behaviors are usually classified into 3 categories in peerplay scenarios (“Solitary Play”, “Parallel Play” and “Group Play”) based on sounds and two main visual cues: the proximity of children and the attention distribution of children. Since the proximity of children can be calculated by the locations, the computation of attention distribution and how to extract meaningful feature descriptions are two important challenges when analyzing children’s social behavior by computer vision methods. In order to solve these challenges, this extended abstract presents a novel method to label children’s social behavior automatically in peer-play scenarios based on visual attention computation.