Low-Rank Embedding for Robust Image Feature Extraction

Robustness to noises, outliers, and corruptions is an important issue in linear dimensionality reduction. Since the sample-specific corruptions and outliers exist, the class-special structure or the local geometric structure is destroyed, and thus, many existing methods, including the popular manifo...

Full description

Saved in:
Bibliographic Details
Published in:IEEE transactions on image processing 2017-06, Vol.26 (6), p.2905-2917
Main Authors: Wong, Wai Keung, Lai, Zhihui, Wen, Jiajun, Fang, Xiaozhao, Lu, Yuwu
Format: Article
Language:English
Subjects:
Algorithm design and analysis
Eigenvalues and eigenfunctions
Feature extraction
image feature extraction
Image reconstruction
low rank representation
Manifolds
Principal component analysis
Robust linear dimensionality reduction
Robustness
subspace learning
Citations: Items that this one cites
Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:Robustness to noises, outliers, and corruptions is an important issue in linear dimensionality reduction. Since the sample-specific corruptions and outliers exist, the class-special structure or the local geometric structure is destroyed, and thus, many existing methods, including the popular manifold learning- based linear dimensionality methods, fail to achieve good performance in recognition tasks. In this paper, we focus on the unsupervised robust linear dimensionality reduction on corrupted data by introducing the robust low-rank representation (LRR). Thus, a robust linear dimensionality reduction technique termed low-rank embedding (LRE) is proposed in this paper, which provides a robust image representation to uncover the potential relationship among the images to reduce the negative influence from the occlusion and corruption so as to enhance the algorithm's robustness in image feature extraction. LRE searches the optimal LRR and optimal subspace simultaneously. The model of LRE can be solved by alternatively iterating the argument Lagrangian multiplier method and the eigendecomposition. The theoretical analysis, including convergence analysis and computational complexity, of the algorithms is presented. Experiments on some well-known databases with different corruptions show that LRE is superior to the previous methods of feature extraction, and therefore, it indicates the robustness of the proposed method. The code of this paper can be downloaded from http://www.scholat.com/laizhihui.
ISSN:1057-7149
1941-0042
DOI:10.1109/TIP.2017.2691543