V-Authenticate: Voice Authentication System for Electorates Living with Disability

Abstract

Every ethnic, religious, and gender communities have electorates living with disability. They represent approximately 0.15 percent of the world’s population roughly one out of every seven. When appropriate mechanism is not in place for such large volume of populace to fully participate in the electoral process, it deters democracy from giving this subset of populace a choice from how they wish to be governed. Electorate voice biometric modality could serve as credential to recognize legitimate voters. The voice recognition process relies on features influenced by the physical attribute of vocal tract and the behavioral features of the individual. Voice biometrics differs from other biometric techniques, in that speech samples are captured dynamically over a period of time. Incorporating technologies like the voice recognition system for authentication for the disabled is a huge step in increasing trust and inclusive participation in the democratic process. In this paper, we present V-authenticate, a voice authentication system to address the issue of valid voter’s recognition and verification for the disabled electorate using their voice trait as biometric input parameter. The developed system was evaluated objectively using Mean Square Error (MSE) and Peak Signal to Noise Ratio (PSNR) qualitative metrics. The result of analysis system performance showed an average MSE value of 0.7 and PSNR was 42.9214 decibel showing the consistency of the applied algorithms to authenticate valid electorate living with disability. The system can be adapted by voting authority to enable electorate living with disability to participate in future electronic democratic decision making.