|
Abstract:
|
Biometric recognition refers to the automatic recognition of individuals based on their physiological or behavioral traits (e.g., fingerprint, face), thus allowing us to confirm or establish identity based on "who he is", rather than by "what he possesses (ID card) or remembers" (password). Despite better security, biometric systems have to contend with problems related to non-universality of the trait, limited degrees of freedom, large intra-class variability and spoof attacks. Although the first Automatic Fingerprint Identification System (AFIS) was installed around 1965, biometric recognition remains a difficult pattern recognition problem. In this talk, we present an overview of biometrics, its advantages and limitation, state-of-the-art performance figures and highlight the role of Statistics in addressing challenges in dealing with accuracy, individuality, and security issues.
|
