Joint training method of sub-dictionaries in multiple characteristic spaces and for face recognition

Abstract

The invention provides a joint training method of sub-dictionaries in multiple characteristic spaces and for face recognition. The joint training method of the sub-dictionaries in the multiple characteristic spaces and for face recognition comprises the steps that original training samples{X[1] to X[N]} are projected in the Eigenface characteristic space, the Laplacianface characteristic space and the Gabor characteristic space to form the sub-dictionary OE, the sub-dictionary OL and the sub-dictionary OG respectively, and joint optimization training is carried on the three sub-dictionaries by using a genetic algorithm to obtain the sub-dictionary NE, the sub-dictionary NL and the sub-dictionary NG. According to the joint training method of the sub-dictionaries in the multiple characteristic spaces and for face recognition, a sample, with the highest distinguishing capacity, in each sub-dictionary can be selected, and thus accuracy of face recognition on the basis of a sparse classifier using the sub-dictionaries can be improved.

Description

technical field [0001] The invention belongs to the technical field of image recognition and relates to a multi-dictionary joint training method. Background technique [0002] Face recognition has always been a popular subject in the field of computer vision. Wright et al. proposed to use the sparse classifier (SRC) based on compressed sensing theory to recognize faces, and achieved good results. However, the algorithm directly uses the training picture as a sparse representation of the L1 norm constraint on the detection picture, which obviously cannot fully represent the characteristics of the face picture to be tested, and the high number of atoms in the dictionary increases the complexity of the encoding. [0003] Therefore, how to learn the optimal dictionary from the original training samples has become a hot research direction. There are currently many dictionary learning algorithms for face recognition: [0004] 1. Metaface, KSVD, etc. all learn the original traini...

AI Technical Summary

Problems solved by technology

[0006] The two types of dictionary learning methods mentioned above are all learning the original training pictures in a single feature space, which cannot make full use of the characteristics of the original training samples, and the generalization ability is not good.

Images