DocumentCode :
2508751
Title :
Exploiting Visual Quasi-periodicity for Automated Chewing Event Detection Using Active Appearance Models and Support Vector Machines
Author :
Cadavid, Steven ; Abdel-Mottaleb, Mohamed
Author_Institution :
Dept. of Electr. & Comput. Eng., Univ. of Miami, Coral Gables, FL, USA
fYear :
2010
fDate :
23-26 Aug. 2010
Firstpage :
1714
Lastpage :
1717
Abstract :
We present a method that automatically detects chewing events in surveillance video of a subject. Firstly, an Active Appearance Model (AAM) is used to track a subject´s face across the video sequence. It is observed that the variations in the AAM parameters across chewing events demonstrate a distinct periodicity. We utilize this property to discriminate between chewing and non-chewing facial actions such as talking. A feature representation is constructed by applying spectral analysis to a temporal window of model parameter values. The estimated power spectra subsequently undergo non-linear dimensionality reduction via spectral regression. The low-dimensional representations of the power spectra are employed to train a Support Vector Machine (SVM) binary classifier to detect chewing events. Experimental results yielded a cross validated percentage agreement of 93.4%, indicating that the proposed system provides an efficient approach to automated chewing detection.
Keywords :
health care; image classification; image sequences; object detection; regression analysis; spectral analysis; support vector machines; video surveillance; active appearance models; automated chewing event detection; model parameter values temporal window; nonlinear dimensionality reduction; power spectra; spectral analysis; spectral regression; support vector machine binary classifier; surveillance video; video sequence; visual quasiperiodicity; Face; Principal component analysis; Shape; Spectral analysis; Support vector machines; Training; Vectors;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Pattern Recognition (ICPR), 2010 20th International Conference on
Conference_Location :
Istanbul
ISSN :
1051-4651
Print_ISBN :
978-1-4244-7542-1
Type :
conf
DOI :
10.1109/ICPR.2010.424
Filename :
5597478
Link To Document :
بازگشت