Chongyang Zhang, Jingbang Qiu, Shibao Zheng, Xiaokang Yang
{"title":"Cross-Layered Hidden Markov Modeling for Surveillance Event Recognition","authors":"Chongyang Zhang, Jingbang Qiu, Shibao Zheng, Xiaokang Yang","doi":"10.1109/ICMEW.2012.37","DOIUrl":null,"url":null,"abstract":"In this paper, a novel Cross-Layered Hidden Markov Model (CLHMM) is proposed for high accuracy and low complexity Surveillance Event Recognition (SER). Unlike existing Layered HMM (LHMM) whose inferences are limited in adjacent layers, cross-layer inferences are designed in CLHMM to strengthen reasoning efficiency and reduce computational complexity. One Common Feature Particle Set (CFPS) is also developed in CLHMM to offer the model an assembly of pixel level observations, expert knowledge and Baum-Welch algorithm are combined to achieve optimized performance in CLHMM learning. Experimental results on typical surveillance test sequences showed that CLHMM outperforms LHMM in terms of accuracy and computational complexity.","PeriodicalId":385797,"journal":{"name":"2012 IEEE International Conference on Multimedia and Expo Workshops","volume":null,"pages":null},"PeriodicalIF":0.0000,"publicationDate":"2012-07-09","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"3","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"2012 IEEE International Conference on Multimedia and Expo Workshops","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/ICMEW.2012.37","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 3
Abstract
In this paper, a novel Cross-Layered Hidden Markov Model (CLHMM) is proposed for high accuracy and low complexity Surveillance Event Recognition (SER). Unlike existing Layered HMM (LHMM) whose inferences are limited in adjacent layers, cross-layer inferences are designed in CLHMM to strengthen reasoning efficiency and reduce computational complexity. One Common Feature Particle Set (CFPS) is also developed in CLHMM to offer the model an assembly of pixel level observations, expert knowledge and Baum-Welch algorithm are combined to achieve optimized performance in CLHMM learning. Experimental results on typical surveillance test sequences showed that CLHMM outperforms LHMM in terms of accuracy and computational complexity.