Advancements in artificial intelligence for biometric system a systematic review
Areej A. Abed, Abdul Monem S. Rahma, Omar A. Dawood · IET conference proceedings. · 2025
Identity recognition, pivotal in artificial intelligence (AI), is identifying and verifying individuals through unique physiological and behavioral traits. This survey examines the shift from unimodal to advanced multimodal systems in identity recognition. Initially, reliance on single biometric indicators presented challenges in accuracy and security, leading to the adoption of multimodal approaches that combine various biometric traits to enhance verification reliability. This survey's key findings include identifying deep learning methodologies as critical to improving verification accuracy, the challenges of integrating multiple biometric modalities, and the ongoing advancements in sensor technology. The survey, covering various applications, from security enhancements to access control, underscores the importance of deep artificial neural networks in crafting these sophisticated systems. This work highlights the significance of these advancements in addressing current limitations in biometric systems and points toward future innovations to enhance the security and efficiency of identity verification. The survey aims to present a thorough overview of the field's current state and prospects through a detailed review of the literature on unimodal and multimodal biometric systems, including methodologies and technological advancements. It addresses the challenges and efficiencies of identity recognition systems, guiding the development of more secure, precise, and user-centric authentication solutions. This comprehensive analysis not only maps the existing research terrain but also anticipates future innovations in security technologies.