Discriminative Dictionary Learning Based on the Double Weighted Constraints
LI Zhengming
YANG Nanyue
Abstract：In order to improve the classification performance of dictionary learning algorithms,a double weighted constraints discriminative dictionary learning algorithm(DWCDL)is proposed. The weighted constraint of atoms is constructed by using the pro?files,and it not only encourages the similar atoms to reconstruct training samples of the same class,but reduces the coherence of at?oms. The weighted constraint of coding coefficients is constructed by using the label information of training samples,and it can en?courage the training sample of the same class to have similar coding coefficients. Experimental results show that the proposed algo?rithm can achieve better classification performance than seven sparse coding and dictionary learning algorithms.
Keywords：dictionary learningweighted constraintimage classification
Publication Date：2017-01-01
Online Publishing Date：2025-08-15(First online date of this platform, not the publication date of the document)
Pages：6( 807-811,843 )