Realtime Anomaly Detection Using Trajectory-Level Crowd Behavior Learning

Aniket Bera, Sujeong Kim, Dinesh Manocha; Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR) Workshops, 2016, pp. 50-57

We present an interactive crowd behavior learning algorithm that can be used for analyzing crowd videos to detect anomalies in realtime for surveillance related applications. Our formulation combines online tracking algorithms from computer vision, non-linear pedestrian motion models from computer graphics, and machine learning techniques to automatically compute the trajectory-level pedestrian behaviors for each agent in the video. These learned behaviors are used to automatically perform motion segmentation to detect anomalous behaviors. We demonstrate the interactive performance using the PETS 2016 ARENA dataset and various indoor and outdoor crowd video benchmarks consisting of tens of human agents.