Face Recognition Dimensionality Reduction Based on LLE and ISOMAP

Conference paper
Part of the Lecture Notes in Electrical Engineering book series (LNEE, volume 208)

Abstract

Local linear embedding (LLE) and isometric feature mapping (ISOMAP) are two basic patterns of nonlinear dimensionality reduction. Their respective strengths and weaknesses in face recognition deserve deep-going comparative study. Therefore, this paper is to test the two patterns’ performance efficiency in different parameters, analyze and summarize the two dimensionality reduction pattern’s characteristics and scope of application, apply LLE and main constituent analysis into face recognition and summarize probability of detection of face recognition.

Keywords

Nonlinear dimensionality reduction Local linear embedding Isometric mapping Face recognition 

References

  1. 1.
    Zejie W (2008) Comparative analysis of two types of nonlinear reduced-dimensional manifold learning algorithm. Academic paper of Shanghai University of Engineering Science 22(1):54–59Google Scholar
  2. 2.
    Ziqiang W, Xu Q, Min K (2008) A summary of manifold learning algorithm. Comput Eng Appl 44(35):9–12Google Scholar
  3. 3.
    Qing L (2008) Semi-supervised handwriting recognition. Chinese Science and Technology University, 49:372–378Google Scholar
  4. 4.
    Varini C, Degenhard A, Nattkemper TW (2006) ISOLLE: LLE with geodesic distance. Neurocomputing 17:1768–1771CrossRefGoogle Scholar
  5. 5.
    Roweis ST, Saul LK (2000) Nonlinear dimensionality reduction by locally linear embedding. Science 8:2323–2326CrossRefGoogle Scholar

Copyright information

© Springer-Verlag London 2013

Authors and Affiliations

  1. 1.Institute of Computer and CommunicationHunan University of TechnologyZhuzhouChina
  2. 2.Hunan City UniversityYiyangChina
  3. 3.Modern Education Technology CenterHunan University of TechnologyZhuzhouChina

Personalised recommendations