Rotational Invariant Dimensionality Reduction Algorithms

Zhihui Lai, Yong Xu, Jian Yang, Linlin Shen, David Zhang

Research output: Journal PublicationArticlepeer-review

93 Citations (Scopus)

Abstract

A common intrinsic limitation of the traditional subspace learning methods is the sensitivity to the outliers and the image variations of the object since they use the L2 norm as the metric. In this paper, a series of methods based on the L21-norm are proposed for linear dimensionality reduction. Since the L21-norm based objective function is robust to the image variations, the proposed algorithms can perform robust image feature extraction for classification. We use different ideas to design different algorithms and obtain a unified rotational invariant (RI) dimensionality reduction framework, which extends the well-known graph embedding algorithm framework to a more generalized form. We provide the comprehensive analyses to show the essential properties of the proposed algorithm framework. This paper indicates that the optimization problems have global optimal solutions when all the orthogonal projections of the data space are computed and used. Experimental results on popular image datasets indicate that the proposed RI dimensionality reduction algorithms can obtain competitive performance compared with the previous L2 norm based subspace learning algorithms.

Original languageEnglish
Article number7502120
Pages (from-to)3733-3746
Number of pages14
JournalIEEE Transactions on Cybernetics
Volume47
Issue number11
DOIs
Publication statusPublished - Nov 2017
Externally publishedYes

Keywords

  • Dimensionality reduction
  • image classification
  • image feature extraction
  • rotational invariant (RI) subspace learning

ASJC Scopus subject areas

  • Software
  • Control and Systems Engineering
  • Information Systems
  • Human-Computer Interaction
  • Computer Science Applications
  • Electrical and Electronic Engineering

Fingerprint

Dive into the research topics of 'Rotational Invariant Dimensionality Reduction Algorithms'. Together they form a unique fingerprint.

Cite this