| Abstrakt: |
Person authentication is the automated process of identifying individuals using computational techniques based on information stored in computer systems. This procedure encompasses critical aspects such as security, robustness, privacy, and prevention of forgery. Traditional biometric systems rely on a single mode of identification, which can fall short in providing high-security levels and are susceptible to noise and exploitation. To address these limitations, we introduce an optimization-enabled, deep learning-based multimodal person authentication system. In this innovative system, we leverage a combination of brainwave signals and fingerprint images to enhance security. To carry out person authentication on both modalities, we employ a Deep Maxout Network (DMN). The output from this network is fused using cosine similarity to yield the final authentication result. An important component of this system is the unique African vultures-Aquila Optimization (AVAO) algorithm, designed to update the weights of the DMN. The AVAO algorithm is constructed by enhancing the African Vulture Optimization Algorithm (AVOA) with the extended exploration capabilities of the Aquila Optimizer (AO). This fusion results in an algorithm that effectively fine-tunes the DMN for optimal performance. Our presented multimodal person authentication system demonstrates outstanding performance, achieving an accuracy of 0.926, sensitivity of 0.940, specificity of 0.928, and an F1-score of 0.921, underscoring its exceptional capabilities. An experimental study also showcases the superior performance of AVAO compared to existing techniques such as Multi-task EEG-based Authentication, Multi-model-based fusion, multi-biometric systems, and Visual secret sharing and super-resolution models, using a variety of metrics. [ABSTRACT FROM AUTHOR] |
