| ECVision indexed and annotated bibliography of cognitive computer vision publications This bibliography was created by Hilary Buxton and Benoit Gaillard, University of Sussex, as part of ECVision Specific Action 8-1 The complete text version of this BibTeX file is available here: ECVision_bibliography.bib
K. Jonsson and J. Kittler and Y. P. Li and J. Matas
ABSTRACT
The paper studies Support Vector Machines (SVMs) in the context of face authentication. Our study supports the hypothesis that the SVM approach is able to extract the relevant discriminatory information from the training data. We believe this is the main reason for its superior performance over benchmark methods. When the representation space already captures and emphasises the discriminatory information content as in the case of Fisher faces, SVMs loose their superiority. SVMs can also cope with illumination changes, provided these are adequately represented in the training data. How ever, on data which has been sanitised by feature extraction (Fisherfaces) and/or normalisation, SVMs can get overtrained, resulting in the loss of the ability to generalise. SVMs involve many parameters and can employ differ ent kernels. This makes the optimisation space rather extensive, without the guarantee that it has been fully explored to find the best solution. |
|
| Site generated on Friday, 06 January 2006
|