Implementation of Voice Biometric System in the Banking Sector

Abstract

Purpose: The voice biometric system for verification and authentication of the user is a more advanced version. The technology is slowly grabbing the attention of researchers and industrialists for customer verification. This method allows users to use a voice instead of any code-based password which can be hacked or forgotten easily. The present research work proposes and describes the voice biometric system that can be implemented in the banking sector. The key of this entire research work is to convert the input voice into a waveform in a multifrequency range and store it in the bank server. During the authentication and verification, a customer will repeat a random phrase given by the voice assistant and then the bank server will match the voice availability.

Design/Methodology/Approach: In this study, we are defining the speech biometric system architecture, which will provide rapid client authentication regardless of language. This technique generates a random word and asks the user to repeat it, rather than asking numerous questions like name, account number details, etc. The banking server will match, recognise, and authenticate the clients' voices by repeating a certain phrase, after which it will determine which customers have access.

Findings/Result: This process will decide the access/denial of the usage of the banking facility to the customer. The overall results conclude that this system will help to enhance the accuracy level of recognition.

Originality/Value: The conceptual framework of the speech biometric system in the banking industry is described in this research study. The system's architecture will assist the banking industry in building a robust voice biometric identification database.

Paper Type: Conceptual Research.