Low-rank representations with incoherent dictionary for face recognition

TL;DR

This paper introduces a semi-supervised low-rank matrix recovery approach for face recognition that learns incoherent dictionaries, improving robustness against variations and noise in images.

Contribution

It proposes a novel method combining low-rank and sparse representations with an incoherent dictionary learning technique for face recognition.

Findings

Effective against illumination, expression, and pose variations

Robust to occlusion and noise in face images

Demonstrated superior performance on multiple face databases

Abstract

Face recognition remains a hot topic in computer vision, and it is challenging to tackle the problem that both the training and testing images are corrupted. In this paper, we propose a novel semi-supervised method based on the theory of the low-rank matrix recovery for face recognition, which can simultaneously learn discriminative low-rank and sparse representations for both training and testing images. To this end, a correlation penalty term is introduced into the formulation of our proposed method to learn an incoherent dictionary. Experimental results on several face image databases demonstrate the effectiveness of our method, i.e., the proposed method is robust to the illumination, expression and pose variations, as well as images with noises such as block occlusion or uniform noises.