ENHANCING BIOMETRIC ACCESS SECURITY IN PHARMACEUTICAL COMPANIES THROUGH CONTACTLESS MULTIMODAL FUSION: A PROPOSED MODEL FRAMEWORK AND SYSTEMATIC REVIEW

Abstract

The recent advancements in contactless multimodal biometric fusion in respect to the pharmaceutical company’s secure access are discussed in the systematic review. The article proposed a new hybrid model framework for secure access. Recent articles that were published between the year 2020 to 2025 were considered using the using the PRISMA approach. From the findings deep learning such as the CNN architectures have shown a decrease in false acceptance rates and in increase in the recognition accuracy (95–99%). However, most models have not been validated in real-world pharmaceutical contexts, are still unimodal, and do not optimize hybrid settings. To bridge these gaps, the suggested Hybrid CNN–Gabor–PSO architecture integrates feature engineering, data augmentation, and hyper parameter tuning. Adaptative learning rate is incorporated in the model after the new feature generation. CNN layers collect spatial patterns, Gabor filters capture frequency and orientation information, and Particle Swarm Optimization (PSO) optimizes the fused feature subset to minimize intra-class variation. In order to guarantee the security of the pharmaceutical companies, the proposed integrated hybrid feature fusion model guarantees secure, reliable, adaptable and effective authentication.