Summary
International Symposium on Nonlinear Theory and its Applications
2010
Session Number:B1L-B
Session:
Number:B1L-B1
Facial Expression Recognition Using a Simplified Head Model and RBF Networks
Koichi Takahashi, Hironobu Fukai, Yasue Mitsukura,
pp.209-212
Publication Date:2010/9/5
Online ISSN:2188-5079
DOI:10.34385/proc.44.B1L-B1
PDF download (395.5KB)
Summary:
In this paper, we propose a facial expression estimation method using radial basis function (RBF) networks. The advantage of an RBF network is that only the arbitrary area of each expression is able to be learned as a classifier. Therefore, we consider whether RBF network is suitable for learning one arbitrary partial expression. The facial expression recognition experiments we conducted demonstrate that the new method estimates the expressions more accurately than the previous method using the Mahalanobis distance.