Facial expression recognition in continuous videos using linear discriminant analysis

In this paper, we address the recognition of facial expressions in continuous videos. We introduce a view- and textureindependent approach that exploits facial action parameters estimated by an appearance-based 3D tracker. We represent the learned facial actions associated with different facial expressions by time series. These time series are then efficiently and compactly represented in Eigenspace and Fisherspace for subsequent recognition. The developed approach is fast and can be used online. Experiments demonstrated the effectiveness of the developed method.