Orthogonal autoencoder regression for image classification

Yang, Zai; Wu, Xinxin; Huang, Pu; Zhang, Fanlong; Wan, Minghua; Lai, Zhihui

doi:10.1016/j.ins.2022.10.068

Cited by 9 publications

(2 citation statements)

References 63 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…They propose to combine the previous projection learning models with linear regression into one objective function and develop one-stage methods. The typical ones include sparse principal component regression (SPCR) [31], joint SPCR (JSPCR) [32], orthogonal autoencoder regression (OAR) [33] etc. To simultaneously take the local geometrical structure and discriminative information into account, Lai et al [34] imposed the L 2,1 -norm on the projection and LPP term to extract robust and discriminative features and proposed a generalised robust regression (GRR) method.…”

Section: Introductionmentioning

confidence: 99%

Low‐rank preserving embedding regression for robust image feature extraction

Zhang

Long

Deng

et al. 2023

IET Computer Vision

View full text Add to dashboard Cite

Although low‐rank representation (LRR)‐based subspace learning has been widely applied for feature extraction in computer vision, how to enhance the discriminability of the low‐dimensional features extracted by LRR based subspace learning methods is still a problem that needs to be further investigated. Therefore, this paper proposes a novel low‐rank preserving embedding regression (LRPER) method by integrating LRR, linear regression, and projection learning into a unified framework. In LRPER, LRR can reveal the underlying structure information to strengthen the robustness of projection learning. The robust metric L2,1‐norm is employed to measure the low‐rank reconstruction error and regression loss for moulding the noise and occlusions. An embedding regression is proposed to make full use of the prior information for improving the discriminability of the learned projection. In addition, an alternative iteration algorithm is designed to optimise the proposed model, and the computational complexity of the optimisation algorithm is briefly analysed. The convergence of the optimisation algorithm is theoretically and numerically studied. At last, extensive experiments on four types of image datasets are carried out to demonstrate the effectiveness of LRPER, and the experimental results demonstrate that LRPER performs better than some state‐of‐the‐art feature extraction methods.

show abstract

Section: Introductionmentioning

confidence: 99%

Low‐rank preserving embedding regression for robust image feature extraction

Zhang

Long

Deng

et al. 2023

IET Computer Vision

View full text Add to dashboard Cite

show abstract

“…With the similar idea, double relaxed regression (DRR) uses two more flexible transformation matrices to address the problem of overfitting well [8]. The orthogonal autoencoder regression (OAR) method integrates relaxation linear regression and orthogonal autoencoder together for improving the flexibility of the label matrix and discrimination of the model [9]. However, the above methods ignore local geometrical structure and the spatial correlations among pixels for the image processing.…”

mentioning

confidence: 99%

Relaxation linear regression with spectral–spatial constrained locality adaptive regularization for hyperspectral image classification

Yang,

Long,

Deng

et al. 2024

Electronics Letters

View full text Add to dashboard Cite

Recently, relaxation linear regression has received increasing attention in image analysis. However, the current relaxation linear regression methods fail to consider the local geometrical structure and spatial information while they are applied for hyperspectral image (HSI) classification. To address the above problems, this letter proposes a novel relaxation linear regression with spectral–spatial constrained locality adaptive regularization (SSLA‐RLR) method for HSI classification. The SSLA‐RLR method not only integrates the locality adaptive graph with relaxation linear regression to adaptively exploit the local geometrical structure for relieving the side effects of noise corruptions, but also takes the spatial correlations between each data point and its neighbours into consideration via spectral–spatial constraint. Finally, extensive experiments are conducted on three benchmark HSI datasets to demonstrate the effectiveness of the proposed method.

show abstract