Search

Sparse representation-based classification (SRC) has become a powerful tool for image recognition. SRC sparsely encodes a test sample over all training samples and then classifies the test sample into the class that generates the minimal reconstruction residual. However, in many real-world applications, nuisances (e.g. illuminations, view directions, pixel corruptions, and occlusion, etc.) may make the representation coefficients of a test sample associated with the training samples from another class greatly larger than those associated with the training samples from the correct class. As a result, the reconstruction residual of the test sample with respect to the other class is smaller than that with respect to the correct class. This inevitably brings a wrong classification of SRC. To address this issue, we propose a sparsity embedding projections (SEP) method, which seeks a low-dimensional embedding subspace where the sparse representation coefficients of a test sample associated with the training samples from the correct class are enlarged, and simultaneously those associated with the training samples from all of the other classes are compressed. Specially, given a training data matrix, SEP tries to find a linear transformation by enhancing the intraclass reconstructive relationship meanwhile suppressing the interclass reconstructive relationship in the low-dimensional embedding subspace. Experimental results on the COIL-20, Extend Yale B, and AR databases show that the proposed method is more effective and robust than other state-of-the-art feature extraction methods with respect to SRC. [ABSTRACT FROM AUTHOR]

Abstract We propose a novel adaptive locality-constrained regularized robust coding (ALRRC) for image recognition by considering both the effects of the test sample's features and the local relationships between the test sample and the training samples in the linear coding procedure. By adaptively calculating out a feature weight matrix of a test sample, ALRRC measures the effects of the test sample's features in the linear coding procedure. Furthermore, ALRRC can obtain the weighted test sample by multiplying the test sample by the feature weight matrix, which has reduced the roles of the aberrant features as far as possible. Similarly, ALRRC can get the weighted training samples by multiplying the training samples by the feature weight matrix. Moreover, using the similarities between the weighted test sample and weighted training samples, ALRRC adaptively calculates out a locality-constrained matrix that can truly characterize the local relationships between the test sample and the training samples. Finally, by incorporating the feature weight matrix and locality-constrained matrix into the linear coding framework, ALRRC is inclined to select the true local training samples to represent the test sample. In addition, we also propose an iterative algorithm to solve the optimization problem of ALRRC. Experimental results on several image databases show that ALRRC is more effective and robust than state-of-the-art linear coding-based methods. [ABSTRACT FROM AUTHOR]