Share Email Print

Proceedings Paper

Anomaly detection driven active learning for identifying suspicious tracks and events in WAMI video
Author(s): David J. Miller; Aditya Natraj; Ryler Hockenbury; Katherine Dunn; Michael Sheffler; Kevin Sullivan
Format Member Price Non-Member Price
PDF $14.40 $18.00

Paper Abstract

We describe a comprehensive system for learning to identify suspicious vehicle tracks from wide-area motion (WAMI) video. First, since the road network for the scene of interest is assumed unknown, agglomerative hierarchical clustering is applied to all spatial vehicle measurements, resulting in spatial cells that largely capture individual road segments. Next, for each track, both at the cell (speed, acceleration, azimuth) and track (range, total distance, duration) levels, extreme value feature statistics are both computed and aggregated, to form summary (p-value based) anomaly statistics for each track. Here, to fairly evaluate tracks that travel across different numbers of spatial cells, for each cell-level feature type, a single (most extreme) statistic is chosen, over all cells traveled. Finally, a novel active learning paradigm, applied to a (logistic regression) track classifier, is invoked to learn to distinguish suspicious from merely anomalous tracks, starting from anomaly-ranked track prioritization, with ground-truth labeling by a human operator. This system has been applied to WAMI video data (ARGUS), with the tracks automatically extracted by a system developed in-house at Toyon Research Corporation. Our system gives promising preliminary results in highly ranking as suspicious aerial vehicles, dismounts, and traffic violators, and in learning which features are most indicative of suspicious tracks.

Paper Details

Date Published: 3 May 2012
PDF: 8 pages
Proc. SPIE 8402, Evolutionary and Bio-Inspired Computation: Theory and Applications VI, 840207 (3 May 2012); doi: 10.1117/12.921476
Show Author Affiliations
David J. Miller, The Pennsylvania State Univ. (United States)
Aditya Natraj, The Pennsylvania State Univ. (United States)
Ryler Hockenbury, The Pennsylvania State Univ. (United States)
Katherine Dunn, Toyon Research Corp. (United States)
Michael Sheffler, Toyon Research Corp. (United States)
Kevin Sullivan, Toyon Research Corp. (United States)

Published in SPIE Proceedings Vol. 8402:
Evolutionary and Bio-Inspired Computation: Theory and Applications VI
Olga Mendoza-Schrock; Mateen M. Rizki; Todd V. Rovito, Editor(s)

© SPIE. Terms of Use
Back to Top