The design and evaluation of adaptive biometric authentication systems: Current status, challenges and future direction

Biometric authentication systems may suffer from decreasing recognition performance due to varying environmental conditions and sample ageing, which cause intra-class variability. Adaptive biometric authentication systems have been proposed to address these deficiencies by dynamically changing their sampling and recognition processes in response to changes in the operating environment. This paper provides the complete discussion on the design of an adaptive biometric authentication system through a systematic review of the existing literature to date. The review shows that further investigation is required to target mobile and/or wearable devices as well as to consider a continuous authentication methodology. In addition to this, the evaluation of adaptive biometric authentications with large-scale datasets is required to validate the feasibility of the system and its scalability in real-world with a comprehensive study of evaluation metrics.