Gabor Feature-Based Fast Neighborhood Component Analysis for Face Recognition

Subspace methods have been very successful in face recognition. Neighborhood components analysis (NCA), one popular subspace method, however, cannot outperform discriminative common vectors (DCV) when applied to face recognition. In this paper, we proposed a Gabor feature-based fast NCA method (Gabor-FNCA). First, we extract multi-scale and multi-orientation Gabor features for more robust and enhanced face recognition. Then, we claimed that the FNCA learning problem would be ill-posed for high dimensional data dimensionality reduction. To address this problem, we first use principal component analysis (PCA) to transform the data in a low-dimensional subspace, and then use the FNCA model which including a Frobenius norm regularizer to learn the linear projection matrix. Experimental results on the ORL and FERET face datasets shows that the proposed Gabor-FNCA method is effective for face recognition.