Arnon Amir, Michael Lindenbaum
IEEE Transactions on Pattern Analysis and Machine Intelligence
Currently there is much interest in the use of biometrics for authentication and identification applications. This has been heightened most recently because of the threat of terrorism. Biometrics authentication and identification systems offer several advantages over systems based on knowledge or possession such as unsupervised (legacy) password/PIN-based systems and supervised (legacy) passport-based systems. To optimize security it is important that biometrics authentication systems are designed to withstand different sources of attack. We identify some such threats to biometrics systems and detail issues related to the tradeoff between security and convenience. We further show how to estimate a biometrics' intrinsic security, sometimes called a biometrics' individuality, with fingerprints as an example. © 2003 Elsevier Science B.V. All rights reserved.
Arnon Amir, Michael Lindenbaum
IEEE Transactions on Pattern Analysis and Machine Intelligence
Fearghal O'Donncha, Albert Akhriev, et al.
Big Data 2021
Bing Zhang, Mikio Takeuchi, et al.
NAACL 2025
Els van Herreweghen, Uta Wille
USENIX Workshop on Smartcard Technology 1999