Optimized gravitational search algorithm for feature fusion in a multimodal biometric system

Abstract

In recent years, multimodal biometric systems have gained significant attention due to their capacity to enhance recognition accuracy and robustness. The integration of multiple biometric traits, such as face, fingerprint, iris, and voice, has shown promising results in addressing the limitations of unimodal systems. However, achieving efficient and accurate feature fusion remains a critical challenge in the development of multimodal biometric systems. This study proposes an optimized approach utilizing the Gravitational Search Algorithm (GSA) for feature fusion in a multimodal biometric system. The objective is to enhance recognition performance by effectively combining complementary information from multiple biometric traits. The evaluation performance to determine the effect of the optimization at threshold values of 0.22, 0.35, 0.5, 0.8 and 1.0 was compared with the traditional GSA and presented. This study achieved the highest GWGSA accuracy at 0.8 and 1.0 thresholds to outperform other thresholds as shown in T able 1. The proposed approach is tested with real-world datasets and compared against existing fusion techniques, demonstrating its superiority in terms of recognition accuracy and computational efficiency.