Abstract:Sparse representation based classification (SRC) has recently been proposed for robust face recognition. To deal with occlusion, SRC introduces an identity matrix as an occlusion dictionary on the assumption that the occlusion has sparse representation in this dictionary. However, the results show that SRC's use of this occlusion dictionary is not nearly as robust to large occlusion as it is to random pixel corruption. In addition, the identity matrix renders the expanded dictionary large, which results in expensive computation. In this paper, we present a novel method, namely structured sparse representation based classification (SSRC), for face recognition with occlusion. A novel structured dictionary learning method is proposed to learn an occlusion dictionary from the data instead of an identity matrix. Specifically, a mutual incoherence of dictionaries regularization term is incorporated into the dictionary learning objective function which encourages the occlusion dictionary to be as independent as possible of the training sample dictionary. So that the occlusion can then be sparsely represented by the linear combination of the atoms from the learned occlusion dictionary and effectively separated from the occluded face image. The classification can thus be efficiently carried out on the recovered non-occluded face images and the size of the expanded dictionary is also much smaller than that used in SRC. The extensive experiments demonstrate that the proposed method achieves better results than the existing sparse representation based face recognition methods, especially in dealing with large region contiguous occlusion and severe illumination variation, while the computational cost is much lower.

Occlusion Robust Face Recognition with Scattering Operator in Gradient Domain

Local Spatial Continuity Steered Sparse Representation for Occluded Face Recognition

Robust spectral regression for face recognition

Towards Mask-robust Face Recognition.

Robust Face Recognition by Fusion Local Singular Value Feature and Deformable Model

Face recognition with contiguous occlusion using linear regression and level set method

Face Recognition under Varying Illumination Based on Gradientface and Local Features

Robust Face Recognition With Occlusion By Fusing Image Gradient Orientations With Markov Random Fields

Robust surface reconstruction from gradient fields

Face Recognition Using Scattering Convolutional Network

Robust 3D Face Recognition by Local Shape Difference Boosting

Robust Feature Extracted by the Generalized Optimal Discriminant Vectors for Face Recognition

Face recognition with the robust feature extracted by the generalized Foley-Sammon transform

Robust Face Recognition Via Gradient-Based Sparse Representation.

Robust Face Recognition Via Occlusion Dictionary Learning.

Facial Expression Recognition with Geometric Scattering on 3D Point Clouds

Face Recognition Using Training Data with Artificial Occlusions.

An efficient method for occluded face recognition

[Sport fishing and animal welfare].

Downsampling Sparse Representation and Discriminant Information Aided Occluded Face Recognition

Robust Face Recognition Via Sparse Representation