Analysis of SubspaceLDA Based for Face Recognition

In this paper, the subspace LDA algorithm for face recognition has been analyzed. This algorithm is evaluated experimentally on ORL database using MATLAB. Analysis evaluates that as the number of training images per individual increases, success rate also goes on increasing but it also causes increase in processing time. When the training images per individual are 5 or 6, it gives maximum success rate with optimized performance time. There is a proportionately high success rate when the eigenface space's dimension is small (40- 60) and comparatively less success rate when eigenface space's dimension is large (180-200). When only significant eigen vectors are used in subspace LDA, then it gives maximum success rate.

[1]  Johannes Stallkamp,et al.  Face Recognition for Smart Interactions , 2007, 2007 IEEE International Conference on Multimedia and Expo.

[2]  Alex Pentland,et al.  Face recognition using view-based and modular eigenspaces , 1994, Optics & Photonics.

[3]  Alex Pentland,et al.  View-based and modular eigenspaces for face recognition , 1994, 1994 Proceedings of IEEE Conference on Computer Vision and Pattern Recognition.

[4]  R. Fisher.  THE USE OF MULTIPLE MEASUREMENTS IN TAXONOMIC PROBLEMS , 1936 .

[5]  Rama Chellappa,et al.  Face recognition using discriminant eigenvectors , 1996, 1996 IEEE International Conference on Acoustics, Speech, and Signal Processing Conference Proceedings.

[6]  Lawrence Sirovich,et al.  Application of the Karhunen-Loeve Procedure for the Characterization of Human Faces , 1990, IEEE Trans. Pattern Anal. Mach. Intell..

[7]  Hyeonjoon Moon,et al.  A verification protocol and statistical performance analysis for face recognition algorithms , 1998, Proceedings. 1998 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (Cat. No.98CB36231).