Aims

Biometrics system concepts is an application-driven course of artificial intelligence on different techniques in identifying/authenticating an individual in an automated, reliable, and fast way using unique physiological (e.g., face, fingerprint, hand, iris, and ear) or behavioral (e.g., keystroke, signature, gait, and speech) characteristics. It introduces concepts, methods, and tools in the field of biometrics and familiarizes learners with current research directions, whilst providing a critical attitude towards its premises and limitations. After following this course:

Learners can explain the added value of biometrics compared to password or token-based authentication, list main biometric applications and compare different types of identity tasks.

Learners can outline the general architecture of biometric systems, discuss biometric system requirements and examine the accuracy or performance of biometric systems.

For the following specific biometric identifiers: face, fingerprint, iris, retina, hand, ear, signature, and keystroke dynamics, learners can restate the typical features used to identify, list acquisition hardware, explain feature extraction techniques and evaluate their strengths and weaknesses.

Learners can construct phyton-based implementations to investigate the performance of biometrics systems.

For one or more biometric identifiers learners can construct phyton-based implementations to train a biometric system. Subsequently learners can justify and assess algorithmic and training design choices.

For a series of listed topics related to biometric systems (e.g., multimodal biometrics, spoofing, ethical and legal implications, specific implementations, and other undiscussed identifiers) learners can assess a paper from the recent international scientific literature.

Finally, learners can assess biometric systems with a technical critical mind and an awareness of the ethical and legal implications of biometric systems.