New Publication: Cancelable Biometric Authentication for IoT Security

Mahmoud Nasr (PhD Student in Computational Neuroscience), Krzysztof Brzostowski, Adam Piórkowski, Fathi E. Abd El-Samie

We are pleased to announce the publication of a new scientific article co-authored by Mahmoud Nasr from the Computational Neuroscience group at Sano Centre for Computational Medicine, together with Krzysztof Brzostowski, Adam Piórkowski, and Fathi E. Abd El-Samie.

The article explores an innovative method for enhancing biometric security in Internet of Things (IoT) applications. It introduces a technique for generating cancelable biometric templates that are both secure and privacy-preserving. By applying Empirical Mode Decomposition (EMD) to deconstruct biometric data into intrinsic components and then encoding them using quaternion mathematics, the method ensures that the resulting templates are non-invertible and highly resistant to spoofing attacks.

The proposed system is not only robust—achieving an Area Under Curve (AUC) of 0.9997 and a near-zero Equal Error Rate—but also lightweight in terms of computational cost, making it particularly well-suited for devices with limited processing power. These results underscore the method’s potential to address critical challenges in biometric security for IoT systems.

This work reflects Sano’s ongoing commitment to advancing secure, efficient, and privacy-focused technologies at the intersection of biometrics, AI, and computational neuroscience.