Mutual Neighborhood Based Discriminant Projection for Face Recognition
Abstract
Linear Discriminant Analysis is optimal under the assumption that the covariance matrices of the conditional densities are normal and all identical. However, this doesn’t hold for many real world applications, such as Facial Image Recognition, in which data are typically under-sampled and non-Gaussian. To address this deficiency the Non-Parametric Discriminant method has been developed, but it requires model selection to be carried out for selecting the free control parameters, making it not easy for use in practice. We proposed a method, Mutual Neighborhood based Discriminant Projection, to overcome this problem. MNDP identifies the samples that contribute most to the Baysesian errors and highlights them for optimization. It is more convenient for use than NDA and avoids the singularity problem of LDA. On facial image datasets MNDP is shown to outperform Eigenfaces and Fisherfaces under various experimental conditions.
Keywords
k-nearest neighbors mutual neighborhood discriminant projection face recognitionReferences
- 1.Fukunaga, K.: Introduction to Statistical Pattern Recognition. Academic Press, Boston (1990)MATHGoogle Scholar
- 2.Fukunaga, K., Mantock, J.: Nonparametric Discriminant Analysis. IEEE Trans. Pattern Anal. Machine Intell. 5, 671–678 (1983)MATHCrossRefGoogle Scholar
- 3.He, X., Cai, D., Yan, S., Zhang, H.J.: Neighborhood Preserving Embedding. In: Proceedings of the Tenth IEEE International Conference on Computer Vision, pp. 1208–1213 (2005)Google Scholar
- 4.Hu, D., Feng, G., Zhou, Z.: Rapid and brief communication: Two-dimensional Locality Preserving Projections (2DLPP) with its Application to Palmprint Recognition. Pattern Recognition 40(1), 339–342 (2007)MATHCrossRefGoogle Scholar
- 5.Martinez, A., Benavente, R.: The AR Face Database. CVC Technical Report, no. 24 (June 1998)Google Scholar