Fuzzy face vault: How to implement fuzzy vault with weighted features

7Citations
Citations of this article
5Readers
Mendeley users who have this article in their library.

This article is free to access.

Abstract

Ari Juel et al from ISIT 2002 presented a fuzzy vault scheme, which is a framework to encrypt a cryptographic key with a fuzzy key [5]. Following their framework, many trials to implement mainly a fuzzy finger vault have been proposed that enables us to keep secretly a key with our finger print. Our work is to focus on instantiating the fuzzy information of the fuzzy vault scheme with human faces instead of fingers. Most of face authentication algorithms are dependent upon weighted features, which are incompatible with the original fuzzy vault scheme. To reflect the level of importance of individual features from feature set, we introduce another layer between captured feature set and points in the polynomial to be interpolated. © Springer-Verlag Berlin Heidelberg 2007.

Cite

CITATION STYLE

APA

Nyang, D. H., & Lee, K. H. (2007). Fuzzy face vault: How to implement fuzzy vault with weighted features. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 4554 LNCS, pp. 491–496). Springer Verlag. https://doi.org/10.1007/978-3-540-73279-2_55

Register to see more suggestions

Mendeley helps you to discover research relevant for your work.

Already have an account?

Save time finding and organizing research with Mendeley

Sign up for free