University of California, Riverside

Department of Electrical and Computer Engineering

Real-Time Systems for Dynamic Activity Understanding, Analysis, and Prediction

Real-Time Systems for Dynamic Activity Understanding, Analysis, and Prediction

Real-Time Systems for Dynamic Activity Understanding, Analysis, and Prediction

February 24, 2014 - 11:10 am
Winston Chung Hall, 205/206


A core goal for intelligent systems is to provide high-level semantic understanding of is environment from raw sensor data.  The challenge is to provide computationally efficient algorithms to perform this analysis in real-time and have the ability to scale to larger environments.  Widespread use of video cameras for monitoring provides semantically rich data streams that can be effectively mined.  This talk presents a hierarchical learning framework for live video analysis to describe behavior in an unsupervised fashion based on object tracking.  Activity models are constructed in an unsupervised fashion from recurrent motion patterns through trajectory clustering and are utilized for real-time characterization and prediction of future activities as well as the detection of abnormalities.  Evaluation on various transportation and surveillance datasets demonstrates the efficacy and utility of the activity analysis framework.  The trajectory framework is the core of an integrated traffic management system that monitors both highways and arterials.  The CalSentry system provides traditional operational traffic measures, roadway usage by vehicle type, and emission estimation on highways and an intersection system delivers continuous turning movement counts and queue analysis.  the above fundamental questions and validates the theoretical conclusions numerically and experimentally.


 Dr. Brendan Morris received his B.S. degree from the University of California, Berkeley (2002) and his Ph.D. degree from the University of California, San Diego (2010). He is an Assistant Professor in Electrical and Computer Engineering and founder of the Real-Time Intelligent Systems Laboratory at the University of Nevada, Las Vegas.  He and his team work on research in computationally efficient systems which utilize computer vision and machine intelligence for situational awareness and scene understanding.

Prof. Morris' research focus has been in real-time sensing and processing for understanding environments and situations with emphasis in transportation.  His interests include unsupervised machine learning for recognizing and understanding activities, real-time measurement, monitoring, and analysis, and driver assistance and safety systems.  His dissertation research on "Understanding Activity from Trajectory Patterns" was awarded the IEEE ITS Society Best Dissertation Award in 2010.  He serves as an Associate Editor for the IEEE Transactions on Intelligent Transportation Systems (ITS) and ITS Magazine.  He will serve as program chair for the 2014 ITS Conference and 2016 Intelligent Vehicles Workshop.




More in Colloquia

More Information 

General Campus Information

University of California, Riverside
900 University Ave.
Riverside, CA 92521
Tel: (951) 827-1012

Department Information

Electrical and Computer Engineering
Suite 343 Winston Chung Hall
University of California, Riverside
Riverside, CA 92521-0429

Tel: (951) 827-2484
Fax: (951) 827-2425
E-mail: E-mail/Questions