pith. sign in

arxiv: 1508.01158 · v2 · pith:RPOT363Dnew · submitted 2015-08-05 · 💻 cs.CV

Socially Constrained Structural Learning for Groups Detection in Crowd

classification 💻 cs.CV
keywords crowdgroupsalgorithmdetectionsocialstructuraltrajectoriesable
0
0 comments X
read the original abstract

Modern crowd theories agree that collective behavior is the result of the underlying interactions among small groups of individuals. In this work, we propose a novel algorithm for detecting social groups in crowds by means of a Correlation Clustering procedure on people trajectories. The affinity between crowd members is learned through an online formulation of the Structural SVM framework and a set of specifically designed features characterizing both their physical and social identity, inspired by Proxemic theory, Granger causality, DTW and Heat-maps. To adhere to sociological observations, we introduce a loss function (G-MITRE) able to deal with the complexity of evaluating group detection performances. We show our algorithm achieves state-of-the-art results when relying on both ground truth trajectories and tracklets previously extracted by available detector/tracker systems.

This paper has not been read by Pith yet.

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.