Representational oriented component analysis (ROCA) for face recognition with one sample image per training class

Subspace methods such as PCA, LDA, ICA have become a standard tool to perform visual learning and recognition. In this paper we propose representational oriented component analysis (ROCA), an extension of OCA, to perform face recognition when just one sample per training class is available. Several...

Full description

Saved in:
Bibliographic Details
Published in2005 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR'05) Vol. 2; pp. 266 - 273 vol. 2
Main Authors De la Torre, F., Gross, R., Baker, S., Vijaya Kumar, B.V.K.
Format Conference Proceeding
LanguageEnglish
Published IEEE 2005
Subjects
Face detection
Face recognition
Image analysis
Image recognition
Independent component analysis
Linear discriminant analysis
Matched filters
Power system modeling
Principal component analysis
Samarium
Online AccessGet full text

Cover

More Information
Summary:Subspace methods such as PCA, LDA, ICA have become a standard tool to perform visual learning and recognition. In this paper we propose representational oriented component analysis (ROCA), an extension of OCA, to perform face recognition when just one sample per training class is available. Several novelties are introduced in order to improve generalization and efficiency: (1) combining several OCA classifiers based on different image representations of the unique training sample is shown to greatly improve the recognition performance. (2) To improve generalization and to account for small misregistration effect, a learned subspace is added to constrain the OCA solution, (3) a stable/efficient generalized eigenvector algorithm that solves the small size sample problem and avoids overfitting. Preliminary experiments in the FRGC Ver 1.0 dataset show that ROCA outperforms existing linear techniques (PCA, OCA) and some commercial systems.
ISBN:0769523722
9780769523729
ISSN:1063-6919
1063-6919
DOI:10.1109/CVPR.2005.301