Taylor-Grey Rider based Deep Recurrent Neural Network using Feature Level Fusion for Cryptography Enabled Biometric System

The authentication, verification, and identification of user can be done using the biometric traits, such as the face, hand vein, finger knuckle, iris, gait, and so on. Among various traits, finger knuckle, face, and hand vein are the important modalities used to recognize the user identity. Due to factors, like image misalignment or illumination variation, the existing unimodal biometric system degrades the performance of recognition. To solve such issues, cryptography enabled multimodal biometric system is developed in this research using the proposed Taylor-Grey Rider based Deep Recurrent Neural Network (Tay-GR based Deep RNN). The proposed Tay-GR is developed by integrating the Taylor series with the Grey Wolf Optimizer (GWO) and Rider Optimization Algorithm (ROA). By integrating the GWO and ROA with Taylor series, the performance of biometric recognition can be increased by reducing the computational cost and time, which can be optimally achieved by updating the Taylor coefficients with the best solution of search agent of wolves. However, the proposed Tay-GR based Deep RNN obtained better performance in terms of accuracy, sensitivity, and specificity with the values of 95.81%, 97.78%, and 95.10%, respectively.