in Advances in Biometric Person Authentication: International Wokshop on Biometric Recognition Systems, IWBRS 2005, Beijing, China, October 22-23, 2005. Proceedings Editors:  Stan Z. Li, Zhenan Sun, Tieniu Tan, Sharath Pankanti, Gérard Chollet, David Zhang.

ABSTRACT:

We propose two new classifiers, one based on the classical Procrustes analysis and the other on the Moore-Penrose inverse in the context of face recognition. The Procrustes based classifier has recognition rates of 97.5%, 96.19%, 71.40% and 96.22% for the ORL, YALE, GIT and the FERET database respectively. The Moore-Penrose classifier has comparative recognition rates of 98%, 99.04%, 87.40% and 96.22% for the same databases. In addition to these classifiers, we also propose new parameters that are useful for comparing classifiers based on their discriminatory power and not just on their recognition rates. We also compare the performance of our classifiers with the baseline PCA and LDA techniques as well as the recently proposed discriminative common vectors technique for the above face databases.

 

(with P. Rajamannar) Roweis-Saul Classifier for Machine Learning 

in NINETH INTERNATIONAL SYMPOSIUM ON ARTIFICIAL INTELLIGENCE AND MATHEMATICS, 2006, FORT LAUDERDALE, FLORIDA.

ABSTRACT:

In 2000, Saul and Roweis proposed locally linear embedding as a tool for  nonlinear dimensionality reduction. In this paper, we modify the LLE algorithm and formulate it as a classifier  in a manner reminiscent of He et al.  Our experiments with the ORL, YALE, FERET face databases and MNIST handwritten  database show that our classifier has recognition rates of 95.42%, 96.67%,  95.28% and 92.50% respectively, clearly outperforming the baseline PCA and  LDA classifiers as well as the recently proposed Laplacianfaces.  We also point out some relationships between the Roweis-Saul classifier, PCA and LDA.