ABSTRACT
The usage of biometrics is becoming routine in everyday life, as fingerprints are used for unlocking smartphones, selfies are used for online identity proofing, and banks use speaker verification for customer authentication; and machines are now claimed to be better than humans at identifying faces. Biometric usage for human–computer interaction is mostly concerned with biometric recognition, or matching. Biometric recognition can be divided into two categories, one-to-many matching, also called identification, and one-to-one matching, also called verification. The chapter introduces basic concepts related to biometric verification and describes four paradigms commonly used for biometric verification, which make use of biometric templates, statistical models, deep neural networks, and biometric cryptosystems. It discusses biometric security, including zero-effort attacks, presentation attacks and their mitigation, and security architectures. The chapter concludes by suggesting possible avenues for future research in biometric verification.
