A Neural Network Approach to Component versus Holistic Recognition of Facial Expressions in Images

Agus Rahardja
Arcot Sowmya
William H. Wilson
School of Computer Science and Engineering
University of New South Wales
Sydney NSW 2052 Australia
Proc. SPIE Vol. 1607, p. 62-70
Intelligent Robots and Computer Vision X: Algorithms and Techniques,
David P. Casasent, Ed., 1991.


The role of features versus the whole in the learning of human facial expressions is explored. A pyramid-like modular network has been developed to learn and identify hand-drawn facial expressions. Because of the nature of the network architecture, image size becomes less of an issue in network learning. The network exhibits a parallel learning capability which could be used to speed up the training process. An analysis of the hidden units of the network reveals that features are used in learning when there is a commonality of facial features in the training patterns. We have also demonstrated attention focusing in the network by masking off specific areas of the face during testing. Our network model creates a "leaner" representation of the original face object and classification is based on this representation. By including the leaner representation and separate key features in the final training set, we can simulate a coarse-to-fine search method, as in image processing.

Order paper on-line from publisher website


Bill Wilson's contact info
Page last updated:
UNSW's CRICOS Provider No. is 00098G