Structured Sparse Error Coding for Face Recognition With Occlusion

Abstract

Face recognition with occlusion is common in the real world. Inspired by the works of structured sparse representation, we try to explore the structure of the error incurred by occlusion from two aspects: the error morphology and the error distribution. Since human beings recognize the occlusion mainly according to its region shape or profile without knowing accurately what the occlusion is, we argue that the shape of the occlusion is also an important feature. We propose a morphological graph model to describe the morphological structure of the error. Due to the uncertainty of the occlusion, the distribution of the error incurred by occlusion is also uncertain. However, we observe that the unoccluded part and the occluded part of the error measured by the correntropy induced metric follow the exponential distribution, respectively. Incorporating the two aspects of the error structure, we propose the structured sparse error coding for face recognition with occlusion. Our extensive experiments demonstrate that the proposed method is more stable and has higher breakdown point in dealing with the occlusion problems in face recognition as compared to the related state-of-the-art methods, especially for the extreme situation, such as the high level occlusion and the low feature dimension.

On the parallel efficiency and scalability of the correntropy coefficient for image analysis

“…The idea of ITL is to use metrics based on nonparametric estimates of Renyi's quadratic entropy as cost functions for the design of adaptive systems. In the past few years, this concept has been successfully applied in the solution of various engineering problems, e.g., automatic modulation classification [19], prediction method for network traffic [20], adaptive filtering [21], and face recognition [22].…”

“…In recent years, the concept of correntropy has been successfully applied to the solution of many problems related to digital image processing, such as automatic face recognition [5,6], facial recognition with occlusion [7], image recognition using MACE filter (correntropy minimum average correlation energy (CMACE) [8], and classification of sidescan sonar imagery [9]. Nevertheless, depending on the input size, the correntropy can demand a high *Correspondence: aluisio@dca.ufrn.br 1 Department of Computer Engineering and Automation, Lagoa Nova, 59078-970 Natal, Brazil computational cost.…”

Eye-Strip based Person Identification based on Non-Subsampled Contourlet Transform

“…Face recognition from occluded images is a challenging task and requires highly discriminative feature extraction methods to achieve successful biometric identification. Occlusions due to scarf, sunglasses and hair falling on the face are commonly addressed problems in the existing literature [6], [7], [8], and [9]. These approaches consider the commonly occurring occlusions.…”

A hierarchical structure with improved OMP sparse representation used with face recognition

“…Many of the face recognition methods have been reported, such as holistic-based feature recognition [6][7][8][9], matching by structuralbased features [10][11][12] as well as methods based on hybrid features [13][14][15]. The first approach is based on facial geometry characteristics (geometric feature based), but it has low efficiency on identification.…”