Fast and Accurate Active Appearance Models

Matthews, I.

Active Appearance Models (AAMs) are generative, non-rigid, parametric models of a particular visual phenomena. They are most often applied to faces, which are perhaps the most popular class of deformable objects in computer vision. There are two elements to using AAMs: first, how to create the model, and second: given a model, how to fit it to a given image? This keynote will give an overview of the standard methods of model construction, then go into detail on the fast, analytical fitting solutions we have developed.

Cite as: Matthews, I. (2006). Fast and Accurate Active Appearance Models. In Proc. HCSNet Workshop on the Use of Vision in Human-Computer Interaction, (VisHCI 2006), Canberra, Australia. CRPIT, 56. Goecke, R., Robles-Kelly, A. and Caelli, T., Eds. ACS. 3.

(from crpit.com) (local if available)