Robust classification using structured sparse representation

Abstract

In many problems in computer vision, data in multiple classes lie in multiple low-dimensional subspaces of a highdimensional ambient space. However, most of the existing classification methods do not explicitly take this structure into account. In this paper, we consider the problem of classification in the multi-subspace setting using sparse representation techniques. We exploit the fact that the dictionary of all the training data has a block structure where the training data in each class form few blocks of the dictionary. We cast the classification as a structured sparse recovery problem where our goal is to find a representation of a test example that uses the minimum number of blocks from the dictionary. We formulate this problem using two different classes of non-convex optimization programs. We propose convex relaxations for these two non-convex programs and study conditions under which the relaxations are equivalent to the original problems. In addition, we show that the proposed optimization programs can be modified properly to also deal with corrupted data. To evaluate the proposed algorithms, we consider the problem of automatic face recognition. We show that casting the face recognition problem as a structured sparse recovery problem can improve the results of the state-of-the-art face recognition algorithms, especially when we have relatively small number of training data for each class. In particular, we show that the new class of convex programs can improve the state-ofthe-art face recognition results by 10% with only 25% of the training data. In addition, we show that the algorithms are robust to occlusion, corruption, and disguise.

Linear Regression Line based Partial Face Recognition

“…Lin and Ming [1] utilizes the extension of Posterior Union Model (PUM) and similarity measure is resolute amongst the class containing single sample [2] [4,5,6,7,8]. The prime work was to design sparse representation into face recognition area, [9] where sparse representation was utilized to renovate and rebuild occluded or blemished facial images as well as to align testing face images to gallery images.…”

Gabor feature based robust representation and classification for face recognition with Gabor occlusion dictionary

“…Such a big occlusion matrix makes the sparse coding process very computationally expensive, and even prohibitive. These two issues are not fully solved by the sparsity based FR improvers [28][29][30][31][32][33][34] [40][41][42]. For instance, only holistic features are considered in [29][30][31][32][33][34][ [40][41][42], FR with occlusion is ignored in [28][32] [33], and no occlusion dictionary is learned in [40][41][42].…”

“…Elhamifar et al [30] discussed classification using structure sparse representation to exploit the block structure of the dictionary, and Yang et al [31] proposed a robust sparse coding model with a maximum likelihood estimator like fidelity term. Moreover, learning a discriminative dictionary under the sparse representation framework for classification has also attracted much attention.…”

Efficient Misalignment-Robust Representation for Real-Time Face Recognition

“…The other emphasizes on proposing new frameworks for (semi) controlled scenarios and with cooperative subjects, which have extensive applications including access control, computer systems, automobiles or automatic teller machines, etc [7]. The goal of the latter branch is for high robustness and high accuracy, and many state-of-the-art works [7], [8], [9], [10], [11], [12], [13], [14] have been proposed along this line to address various challenges, including face corruption, occlusion, misalignment and the variations of illumination, expression, etc.…”