Pseudo-Entropy Similarity for Human Biometrics
Abstract
With complex multimedia data, we see the emergence of biometric identification/authentication systems in which the fundamental operation is the similarity assessment of natural information-carrying objects. We have developed a similarity measure JeK , based on new notion of pseudo-entropy . The measure exhibits several features that match experimental findings in multimedia perception. We show how the measure can be used for identification/authentication of complex biometric objects, such as faces and its emotions, voices, and so on. We address the use of the pseudo-entropy measure JeK to deal with relations among the varied properties of 1D, 2D and 3D biometric objects.
Cite
Text
Kompanets et al. "Pseudo-Entropy Similarity for Human Biometrics." European Conference on Computer Vision, 2002. doi:10.1007/3-540-47917-1_8Markdown
[Kompanets et al. "Pseudo-Entropy Similarity for Human Biometrics." European Conference on Computer Vision, 2002.](https://mlanthology.org/eccv/2002/kompanets2002eccv-pseudo/) doi:10.1007/3-540-47917-1_8BibTeX
@inproceedings{kompanets2002eccv-pseudo,
title = {{Pseudo-Entropy Similarity for Human Biometrics}},
author = {Kompanets, Leonid and Bobulski, Janusz and Wyrzykowski, Roman},
booktitle = {European Conference on Computer Vision},
year = {2002},
pages = {67-77},
doi = {10.1007/3-540-47917-1_8},
url = {https://mlanthology.org/eccv/2002/kompanets2002eccv-pseudo/}
}