Protecting Video-Based Biometric Systems from Fraud Attacks with Multi-Feature Fusion

Abstract

Spoofing assaults, in which malicious actors employ forged biometric attributes in an effort to trick biometric systems, present a substantial peril to the security of video-based biometric systems. Antispoofing methods are designed to identify and thwart such attacks through the differentiation of authentic and counterfeit biometric characteristics. Multifeature fusion-based approaches have garnered considerable interest in recent times on account of their capacity to integrate data from numerous sources, thereby resulting in enhanced performance. An examination of anti-spoofing methods in video-based biometric systems, with an emphasis on multifunctional fusion strategies, is the subject of this research. paper. The paper provides a comprehensive overview of the difficulties linked to spoofing attacks, analyzes different attributes that can be extracted from videos to thwart spoofing attempts, and proposes a range of fusion techniques to integrate these attributes. Additionally, the article examines the benefits and drawbacks of multifunctional fusion methodologies, underscores contemporary research. patterns, and predicts future developments in this domain.