DocumentCode :
3673952
Title :
A multiple server scheme for fingerprint fuzzy vaults
Author :
Jesse Hartloff;Matthew Morse;Bingsheng Zhang;Thomas Effland;Jennifer Cordaro;Jim Schuler;Sergey Tulyakov;Atri Rudra;Venu Govindaraju
Author_Institution :
University at Buffalo, SUNY, NY, USA
fYear :
2015
fDate :
6/1/2015 12:00:00 AM
Firstpage :
119
Lastpage :
127
Abstract :
In this work, we present a multiple server fingerprint verification scheme that provides enhanced template security by eliminating several known vulnerabilities of the fuzzy vault scheme. We secure templates from adversarial attacks in honest-but-curious server scenarios by utilizing commutative encryption in which the raw fingerprint template is never used in matching or storage. In this system, there is a matching server that performs the enrollment and matching functions on fingerprint data that has been encrypted by a separate encryption server. Since the encrypted template is stored at one server and the encryption key is on another server, an attacker would have to compromise both servers to decrypt the data. Even in this case, the templates are protected by the fuzzy vault scheme. Thus, this scheme limits an attacker´s ability to attack active users even after compromising both servers providing multiple layers of template security.
Keywords :
"Encryption","Servers","Elliptic curves","Polynomials","Decoding"
Publisher :
ieee
Conference_Titel :
Computer Vision and Pattern Recognition Workshops (CVPRW), 2015 IEEE Conference on
Electronic_ISBN :
2160-7516
Type :
conf
DOI :
10.1109/CVPRW.2015.7301327
Filename :
7301327
Link To Document :
بازگشت