On parsing visual sequences with the hidden Markov model
Journal on Image and Video Processing
Statistical motion information extraction and representation for semantic video analysis
IEEE Transactions on Circuits and Systems for Video Technology
Discriminative human action recognition in the learned hierarchical manifold space
Image and Vision Computing
Event detection using multiple event probability sequences
ICIP'09 Proceedings of the 16th IEEE international conference on Image processing
Mining rules to explain activities in videos
CIKM '10 Proceedings of the 19th ACM international conference on Information and knowledge management
Anomaly detection over spatiotemporal object using adaptive piecewise model
PRICAI'10 Proceedings of the 11th Pacific Rim international conference on Trends in artificial intelligence
Finding "unexplained" activities in video
IJCAI'11 Proceedings of the Twenty-Second international joint conference on Artificial Intelligence - Volume Volume Two
Real-Time exact graph matching with application in human action recognition
HBU'12 Proceedings of the Third international conference on Human Behavior Understanding
EEM: evolutionary ensembles model for activity recognition in Smart Homes
Applied Intelligence
Ongoing human action recognition with motion capture
Pattern Recognition
A real-time system for motion retrieval and interpretation
Pattern Recognition Letters
Hi-index | 0.01 |
Changes in motion properties of trajectories provide useful cues for modeling and recognizing human activities. We associate an event with significant changes that are localized in time and space, and represent activities as a sequence of such events. The localized nature of events allows for detection of subtle changes or anomalies in activities. In this paper, we present a probabilistic approach for representing events using the hidden Markov model (HMM) framework. Using trained HMMs for activities, an event probability sequence is computed for every motion trajectory in the training set. It reflects the probability of an event occurring at every time instant. Though the parameters of the trained HMMs depend on viewing direction, the event probability sequences are robust to changes in viewing direction. We describe sufficient conditions for the existence of view invariance. The usefulness of the proposed event representation is illustrated using activity recognition and anomaly detection. Experiments using the indoor University of Central Florida human action dataset, the Carnegie Mellon University Credo Intelligence, Inc., Motion Capture dataset, and the outdoor Transportation Security Administration airport tarmac surveillance dataset show encouraging results.