Robust Face Recognition With Kernelized Locality-Sensitive Group Sparsity Representation

Abstract

Abstract-In this paper, a novel joint sparse representation method is proposed for robust face recognition. We embed both group sparsity and kernelized locality-sensitive constraints into the framework of sparse representation. The group sparsity constraint is designed to utilize the grouped structure information in the training data. The local similarity between test and training data is measured in the kernel space instead of the Euclidian space. As a result, the embedded nonlinear information can be effectively captured, leading to a more discriminative representation. We show that, by integrating the kernelized localsensitivity constraint and the group sparsity constraint, the embedded structure information can be better explored, and significant performance improvement can be achieved. On the one hand, experiments on the ORL, AR, extended Yale B, and LFW data sets verify the superiority of our method. On the other hand, experiments on two unconstrained data sets, the LFW and the IJB-A, show that the utilization of sparsity can improve recognition performance, especially on the data sets with large pose variation.

Boosting the Face Recognition Performance of Ensemble Based LDA for Pose, Non-uniform Illuminations, and Low-Resolution Images

“…However, the authors did not provide the information about the occlusion and the maximum angle for pose variation investigation in their paper. In [16], authors describe a Kernelized Locality-sensitive Group Sparsity Representation (KLS-GSRC) method for face recognition. Extensive experiments were conducted on advanced databases, such as the LFW, the ORL, Extended Yale B, and the AR dataset.…”

Kernel group sparse representation classifier via structural and non-convex constraints

“…) . The ADMM iterates satisfy that With the fact that the closedness and properness of our cost function (13) are clear, we further need to make sure that all the subproblems be convex and have convergent solutions. It is evident that subproblem (17) is strictly convex and with closed-form solution (18) , the remaining issue for us is to prove the convexity and convergence of subproblem (19) .…”

“…Both the theoretical analysis and the experimental results have showed the promising performance of GSC, outperforming both SRC and CRC. More recently, it has been verified that the property of locality preservation is more important for a classifier [13,14] . As a result, many regression-based works have been proposed, such as integrating the data locality into the constraints of l 1 -norm [15,16] , l 2 -norm [17,18] or group norm [19,20] , for improvement.…”

“…However, it still suffers from the norm normalization problem [23] . Some other works resort to kernel trick to convert linear algorithms into nonlinear forms, such as Kernel SRC (KSRC) [24,25] , Kernel CRC (KCRC) [17] and Kernel GSC (KGSC) [13,26,27] . These approaches map the original data into a high-dimensional feature space by using a nonlinear kernel function, and then perform linear optimization in the feature space with the inner products.…”

Fast search real‐time face recognition based on DCT coefficients distribution

“…Jiang et al [17] proposed dictionary decomposition based on a sparse and dense hybrid representation method to solve the problem of corrupted training data and insufficient representative sample for each class. Tan et al [18] proposed group sparsity and kernelised locality‐sensitive method, and the local similarity is measured instead of the Euclidian distance. To reduce the feature vectors, the frequency components of face are extracted by the transformation of discrete cosine transform (DCT) [19–23] and Fourier [24].…”