Generative Artificial Intelligence for Next-Generation Security Paradigms

4Biometric Fusion: Exploring Generative AI Applications in Multi-Modal Security Systems

Suryakanta¹, Ritu²*, Anu Rani³, Neerja Negi⁴, Surya Kant Pal⁵ and Kamalpreet Singh Bhangu²

¹Department of Computer Science & Engineering, Chandigarh University, Punjab, India

²Amity School of Engineering and Technology, Amity University, Punjab, India

³Department of Computer Science & Engineering, SCSET, Bennett University, Greater Noida, India

⁴Department of Computer Application SCA, MRIIRS, Greater Noida, India

⁵Department of Mathematics, SSBSR, Sharda University, Greater Noida, Uttar Pradesh, India

Abstract

The integration of biometric technologies with generative AI promises advances in multi-modal security systems in the quest for increased security and efficiency. This chapter explores the intersection of biometric modalities such as facial recognition, fingerprint analysis, and voice identification with generative AI techniques like neural networks and deep learning algorithms. By combining their capabilities, multi-modal security systems can realize greater accuracy, robustness, and adaptability. The chapter introduces basic ideas of biometric fusion, especially how the integration of different sources of biometric data can overcome the limitations of individual modalities and improve overall performance. Several fusion strategies are introduced: feature-level, score-level, and decision-level fusion and the advantages and disadvantages of each. The chapter then goes into the application ...