Enhancing Person Recognition with Convolutional Neural Networks in Multimodal Biometrics

An identity management system aims to provide safe, secure as well as easy access to information to an authorized owner. The objective of this work is to identify the individual’s unique identity by implementing a variety of private factors such as PINs, passwords, and tokens. These are typically utilized in standard methods, which can have drawbacks like posting and cracking. Biometric authentication is widely adopted in various domains, including physical access control, mobile devices, financial transactions, border security, and government services, due to its accuracy, security, and convenience. It is basically built for rural area people for distributing ration. As their ridges of the finger goes off due to various activities they carry out in farms, which in turn leads to rejection of the fingerprint. So, when it fails, we can use voice as our second authentication. In this work, we focus on biometric authentication and voice-based authentication, called multi modal biometric authentication utilizing convolutional neural networks.