Relevant Component Analysis for static facial expression classification
Matteo Sorci, Giulio Antonini, Jean‐Philippe Thiran · Infoscience (Ecole Polytechnique Fédérale de Lausanne) · 2005
This paper addresses the issue of automatic classification of the six universal emotional categories (joy, surprise, fear, anger, disgust, sadness) in the case of static images. Appearance parameters are extracted by an active appearance model(AAM) representing the input for the classifi- cation step. We introduce Relevant Component Analysis (RCA) in the context of facial expression recognition framework and we test this method against several other classi- fication techniques, including LDA, GDA and SVM, on the Cohn-Kanade database.