In this paper, an automatic system for recognizing expression intensity is proposed. Modified Active Appearance Model (MAAM) is utilized to extract facial feature points (FFPs), and then, according to the FFPs' position, the sequence is preprocessed. Coarse-to-fine pyramid algorithm is employed to track FFPs for extracting 23 optical flow vectors, and eliminating the error caused by rigid movement of head. Expression intensity is recognized by multilayer hybrid classifier. Support Vector Machine (SVM) classifies the expression in the form of optical flow vectors, and KNN classifier recognizes the intensity. We conduct the experiments on Cohn-Kanade expression database and the result shows good effect. © 2013 Springer-Verlag.
CITATION STYLE
Mao, X., Wang, C., & Xue, Y. (2013). Expression intensity recognition based on multilayer hybrid classifier. In Advances in Intelligent Systems and Computing (Vol. 194 AISC, pp. 739–748). Springer Verlag. https://doi.org/10.1007/978-3-642-33932-5_69
Mendeley helps you to discover research relevant for your work.