Boosting for Learning a Similarity Measure in 2DPCA Based Face Recognition
Zhijie Xu, Jianqin Zhang, Xiwu Dai · 2009
In this paper we address the problem of identifying the similarity measure for face recognition. The similarity measure plays an important role in pattern classification. However, with reference to 2D image matrix based methods for face recognition, such as two dimensional principal component analysis (2DPCA), where the features extracted are matrixes instead of single vectors, studies on the similarity measure are quite few. We propose a new method to identify the similarity measure by boosting, which is called boosted similarity measure. Experimental results on two famous face databases show that generally the proposed method outperforms the state of the art methods.