Social interaction discovery by statistical analysis of F-formations

We present a novel approach for detecting social interactions in a crowded sceneby employing solely visual cues. The detection of social interactions in unconstrainedscenarios is a valuable and important task, especially for surveillance purposes. Ourproposal is inspired by the social signaling literature, and in particular it considers thesociological notion of F-formation. An F-formation is a set of possible configurationsin space that people may assume while participating in a social interaction. Our systemtakes as input the positions of the people in a scene and their (head) orientations; then,employing a voting strategy based on the Hough transform, it recognizes F-formationsand the individuals associated with them. Experiments on simulations and real datapromote our idea.