Abstract

Biometric technologies are playing an increasingly central role in critical identity verification systems. However, persistent concerns around trust, fairness, and security limit their widespread adoption. To build trustworthy systems, it is essential to address key pillars such as performance, resilience to presentation attacks, demographic fairness, explainability, and privacy. This talk explores how emerging technologies—including hyperspectral imaging, which fuses imaging with spectroscopy—can enhance both the security and inclusivity of biometric systems. We will also present novel, low-cost solutions for secure smartphone unlocking and data protection. Special attention will be given to the impact of skin tone variations on the effectiveness of presentation attack detection in mobile biometrics, highlighting current efforts to mitigate these biases. Together, these advancements chart a path toward a more inclusive, resilient, and ethically grounded approach to human recognition through trustworthy AI.

About the Speaker

Emanuela Marasco is an Assistant Professor in the Department of Information Sciences and Technology, affiliated with Computer Science at George Mason University. Her research expertise spans cybersecurity, biometrics, machine learning, deep learning, and computer vision. She has contributed to premier conferences such as IEEE WACV, IJCB, ICIP, and BigData and esteemed journals such as ACM Computing Surveys, Wiley, Springer, and PRL. Dr. Marasco has received two NSF EAGER Awards for research in hyperspectral biometrics and has actively collaborated on projects funded by NSF, DOJ.