Blind authentication
Quick and Provably secure authentication without negotiating out of precision. Supports generic classifiers such as Neural Network and SVMs. Useful with a variety of biometric features of fixed length. Ideal for applications such as biometric ATMs, access from public terminals.
A biometric is a physiological or behavioral characteristic of a human being that can distinguish one person from another and that can theoretically be used for identification or verification of identity. Different biometric systems are Physiological Biometry and Behavioral Biometry.
Characteristics of blind authentication
• Strong encryption
• Non-Reputable Authentication
• Protection against repeat and client attacks
• Revocability
Authentication
Blind authentication process: Linear kernel calculation for encrypted entity vectors. At no time, the identity vectors x, w or the intermediate results xi and wi are revealed to anyone.
• Server Security: Hacker gains access to the template database. Hacker is on the database server during authentication. Impostor attempting blind attacks from a remote machine
• Client Security: Hacker gains access to the user's biometric or private key. Passive attack on the user's computer.
• Network security: the attacker gains access to the network.
• Privacy - Concern to disclose personal information The template is never revealed to the server. Concern about being tracked: use different keys for different applications.