Robust face recognition via low-rank sparse representation-based classification

Abstract

Face recognition has attracted great interest due to its importance in many real-world applications. In this paper, we present a novel low-rank sparse representation-based classification (LRSRC) method for robust face recognition. Given a set of test samples, LRSRC seeks the lowest-rank and sparsest representation matrix over all training samples. Since low-rank model can reveal the subspace structures of data while sparsity helps to recognize the data class, the obtained test sample representations are both representative and discriminative. Using the representation vector of a test sample, LRSRC classifies the test sample into the class which generates minimal reconstruction error. Experimental results on several face image databases show the effectiveness and robustness of LRSRC in face image recognition.

Unlabeled data driven cost-sensitive inverse projection sparse representation-based classification with 1/2 regularization

“…This section mainly verifies the convergence, role of unlabeled data, sparsity, effectiveness for class imbalance problem as well as comparison with the latest published results and some deep-learning algorithms. Since the inverse projection sparse representation-based classification methods [21][22][23][24] have been proven to be superior to some classic classifiers, here S 1/2 -L 1/2 -PFSRC-CS is only compared with S 1/2 -L 1/2 -PFSRC, LR-S-PFSRC [24], PFSRC [23], and LRSRC [18]. And L 1/2 -ISRC-CS is only compared with L 1/2 -ISRC, ISSRC [22], IPRC [21], and SRC [16].…”

“…Wright et al [16] proposed sparse representation-based classification (SRC), which was a sparse coding technique based on sufficient labeled samples per class without learning. SRC has made remarkable success in pattern recognition fields, such as face recognition [17,18] and tumor recognition [19,20]. However, 2 Construction of S 1/2 -L 1/2 -PFSR model…”

Score level fusion in representation-based classification method for face recognition

“…Currently, and in the past few years, representation-based classification method (RBCM), 1,2 which can be applied in a wide variety of fields – e.g. face recognition 3 – 6 and hyperspectral imagery classification – has received more attention. 7 – 9 RBCM represents a test sample as a linear combination of training samples and then employs the representation results to classify the test sample.…”