This paper will explore the effectiveness of active appearance models (AAM) at extracting emotional information as well as usefulness of AAMs in the role of emotion classification. This paper examines what has been accomplished recently in this held by reviewing the various types of classification schemes and databases used. Three types of classification techniques are presented: Euclidean Distance Measure, Gaussian Mixture Models, and the Support Vector Machine. Each will be gauged by its effectiveness in the classification of emotional expressions. The technique presented involves the creation of an Active Appearance Models (AAM) trained on face images taken from a publicly available database. This model is a representation of the shape and texture variation of the image, which is key to expression recognition. In each experiment model parameters from the AAM are used as input into a classification scheme, which is used for expression identification. The results of this study will demonstrate the effectiveness of AAMs in capturing the important facial structure for expression identification and also help suggest a framework for future development.

Recommended Citation: Ratliff M., Patterson E., Janicki T., Guinn C., (2010). Active Appearance Models for Affect Recognition using Facial Expressions . UNCW MS CSIS Proceedings. V. 4 , N. 4 .