Nuclear norm-based matrix regression preserving embedding for face recognition

Deng, Yang‐Jun; Li, Heng-Chao; Wang, Qi; Du, Qian

doi:10.1016/j.neucom.2018.05.078

Cited by 15 publications

(14 citation statements)

References 37 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The proposed TDOA positioning method relies on the idea of reweighting [17,19], which has been applied in the well-known iterative reweighted least squares (IRLS) algorithm. Adapting IRLS to the considered TDOA positioning problem, we can express the associated cost function as [29]…”

Section: Methods Of Reweightingmentioning

confidence: 99%

See 1 more Smart Citation

TDOA source positioning in the presence of outliers

Yang²,

Zhang

et al. 2019

IET signal process.

View full text Add to dashboard Cite

Source localisation using time difference of arrival (TDOA) measurements has drawn considerable attention in the past few decades. The presence of outliers in TDOAs could deteriorate the localisation performance significantly. Under the reasonable assumption that outliers are sparse and they do not dominate in the raw measurement set, a computationally efficient outlier-robust TDOA localisation method is proposed in this study. It integrates the half-quadratic minimisation and reweighted least absolute shrinkage and selection operator to iteratively identify the outliers and find the source location estimate using the TDOA inliers only. Both additive and multiplicative forms of the proposed method are established. Simulation results demonstrate the computational efficiency and effectiveness of the developed algorithms.

show abstract

Section: Methods Of Reweightingmentioning

confidence: 99%

“…which is parameterised on the residual error vector v k + 1 (x) evaluated at the candidate source position x. According to the HQ minimisation procedure given in (19), the minimiser of ϕ(v k + 1 (x)) can ideally be found via executing in sequence the following two steps:…”

Section: Source Position Update Using Hq Minimisationmentioning

confidence: 99%

TDOA source positioning in the presence of outliers

Yang²,

Zhang

et al. 2019

IET signal process.

View full text Add to dashboard Cite

show abstract

“…In this section, we demonstrate the effectiveness of the proposed B2DMRPDE method on five datasets: Yale [47], Extended Yale B [48], CMU-PIE [49], AR [50] and LFW [51]. To evaluate the performance, we compare our proposed method with several state-of-the-art dimensionality reduction algorithms such as 2DPCA [8], 2DLDA [10], 2DNPE [16], 2DSPP [17], B2DLPP [19], B2DNPE [22], B2DNPDE [23] and NN-MRPE [41].…”

Section: Methodsmentioning

confidence: 99%

“…More recently, low-rank minimization problems have been extensively researched, and successfully utilized in matrix recovery [24]- [27], data clustering [28], [29], image denoising [30] and recognition [5], [31]- [41]. It has been proven that the nuclear-norm (NN) can be an effective convex surrogate of the rank term in the low-rank minimization problem [42].…”

Section: Introductionmentioning

confidence: 99%

“…These methods achieved superior performance on data containing various types of noises, e.g., occlusion, illumination, pixel corruption, or mixed noise. To further exploit the 2D structure of the image, Deng et al [41] proposed nuclear-normbased bilateral two-dimensional matrix regression preserving embedding (NN-MRPE) for face recognition. However, NN-MRPE is an unsupervised learning method and does not take the label information into account.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Bilateral Two-Dimensional Matrix Regression Preserving Discriminant Embedding for Corrupted Image Recognition

Zhang

Wang

2019

IEEE Access

View full text Add to dashboard Cite

Nuclear-norm-based matrix regression (NMR) methods have been successfully applied for the recognition of corrupted images. However, most of these methods do not consider the label information and are classified as unsupervised learning methods. In this paper, we propose a new regression-based algorithm, named bilateral two-dimensional matrix regression preserving discriminant embedding (B2DMRPDE). The proposed algorithm constructs the within-reconstruction graph and between-reconstruction graph using NMR. Then, B2DMRPDE aims to seek a subspace in which the within-class reconstructive residual is minimized and the between-class reconstructive residual is maximized based on Fisher's criterion. Hence, B2DMRPDE can capture the potential discriminative information for classification. To enhance the classification effectiveness, we present a new NMR-based classifier to determine the class label of the testing sample. Extensive experiments on face image databases were performed, and the results validate the effectiveness of the proposed method. INDEX TERMS Corrupted image, face recognition, low-rank, matrix regression, nuclear-norm.

show abstract

Robust mixed-norm constrained regression with application to face recognitions

Sang

et al. 2020

Neural Comput & Applic

View full text Add to dashboard Cite

Nuclear norm-based matrix regression preserving embedding for face recognition

Cited by 15 publications

References 37 publications

TDOA source positioning in the presence of outliers

TDOA source positioning in the presence of outliers

Bilateral Two-Dimensional Matrix Regression Preserving Discriminant Embedding for Corrupted Image Recognition

Robust mixed-norm constrained regression with application to face recognitions

Contact Info

Product

Resources

About