In this paper, we describe a weighted principal geodesic analysis (WPGA) method to extract features for gender classification based on 2.5D facial surface normals (needle-maps) which can be extracted from 2D intensity images using shape-from-shading (SFS). By incorporating the weight matrix into principal geodesic analysis (PGA), we control the obtained principal axis to be in the direction of the variance on gender information. Experiments show that using WPGA, the leading eigenvectors encode more gender discriminating power than using PGA, and that gender classification based on leading WPGA parameters is more accurate and stable than based on leading PGA parameters. © Springer-Verlag Berlin Heidelberg 2007.
CITATION STYLE
Wu, J., Smith, W. A. P., & Hancock, E. R. (2007). Weighted principal geodesic analysis for facial gender classification. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 4756 LNCS, pp. 331–339). https://doi.org/10.1007/978-3-540-76725-1_35
Mendeley helps you to discover research relevant for your work.