Plenary Lecture
Human Action Recognition in Videos
Professor Mariofanna Milanova
Computer Science Department
University of Arkansas at Little Rock
USA
E-mail: mgmilanova@ualr.edu
Abstract: Human action recognition in videos is a prominent field in multimedia research with numerous fundamental applications, including video surveillance, patient monitoring systems, law enforcement, video indexing and human computer interface. While various algorithms have been developed for human action recognition the topic still remains a great challenge. There are infinite variations that can result from the combination of the various factors including the subject’s body shape, motion, intention, surrounding environment and the dynamic context. Typical scenarios include scenes with cluttered, moving backgrounds, non-stationary camera, scale variations, changes in light and view points, etc. In this talk, novel algorithms for human action recognition are presented. The new algorithms are based on different types of features, such as for example contour-based features or silhouette-based features. For the classification step, experiments are obtained using support vector machine (SVM), K-mean and structure similarity measurement. Public data sets are illustrated and recognition methodologies performance are evaluated and compared. The obtained results demonstrate a promising solution for improved accuracy and efficiency.
Brief Biography of the Speaker: Mariofanna Milanova is a Professor of Computer Science Department at the University of Arkansas at Little Rock since 2001. She received a M.Sc. in Expert Systems and Artificial Intelligence in 1991. She obtained a Ph.D. in Computer Science in 1995 from the Technical University, Sofia, Bulgaria. Dr. Milanova conducted post-doctoral research in visual perception at the University of Paderborn, Germany. Dr. Milanova has extensive academic experience at various academic and research organizations worldwide. Dr. Milanova serves as a book editor for two books and as associate editor for various international journals. Her main research interests are in the areas of artificial intelligence, computer vision and communications, machine learning, privacy and security based on biometric research. She has published and co-authored more than 80 publications, over 53 journal papers, 11 book chapters, numerous conference papers and has 2 patents.