Advertisement

Face Recognition Human–Machine Comparison Under Heavy Lighting

  • Evgeny V. Mozgunov
Conference paper

Abstract

We demonstrate the performance of the Fisherface method for face recognition compared to human eye and simple Eigenface method. These methods do not involve many adjustable parameters. Images undergo the principal component analysis (PCA) and linear discriminant analysis (LDA). The goal of the work is a detailed comparison of the rates of false recognition between the computer vision methods and human perception. We find that humans show more flexibility and perform perfectly on easy tasks, whereas on tasks that are impossible to humans, Fisherface method also fails.

Keywords

Face Recognition Linear Discriminant Analysis Training Image Eigenvalue Equation False Recognition 
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

References

  1. 1.
    Belhumeur N, Hespanha P, Kriegman J (1997) Eigenfaces vs. fisherfaces: recognition using class specific linear projection, pattern analysis and machine intelligence. IEEE Trans 19(7):711Google Scholar
  2. 2.
    Turk M, Pentland A (1991) Eigenfaces for recognition. J Cognit Neurosci Arch 3(1):71–86CrossRefGoogle Scholar
  3. 3.
    Shin H-C, Park JH, Kim S-D (2007) Combination of warping robust elastic graph matching and kernel-based projection discriminant analysis for face recognition. IEEE Trans Multimed 9(6):1125–1136CrossRefGoogle Scholar
  4. 4.
    Gokberk B, Dutagaci H, Ulas A, Akarun L, Sankur B (2008) Representation plurality and fusion for 3-D face recognition, systems, man, and cybernetics, Part B: cybernetics. IEEE Trans 38(1):155–173Google Scholar
  5. 5.
    Georghiades AS, Belhumeur PN, Kriegman DJ (2001) From few to many: illumination cone models for face recognition under variable lighting and pose. IEEE Trans Pattern Anal Mach Intell 23(6):643CrossRefGoogle Scholar

Copyright information

© Springer Japan 2015

Authors and Affiliations

  1. 1.California Institute of TechnologyPasadenaUSA

Personalised recommendations