Authentication of Voter Using Convolutional Neural Network and Blockchain
Kapil Upadhyay, Aman Bansal, Adarsh Gupta, Vipashi Kansal, Upma Jain, Shruti Bhatla · 2023
For elected officials and government policies to be credible and legitimate in today's democracies, voting systems must be fair and honest. Voter impersonation and ballot stuffing are just two examples of the many fraud and manipulation schemes that can be used against traditional voting systems. We suggest a novel method for voter authentication that combines blockchain technology and convolutional neural networks (CNNs) to address these issues. Our suggested method entails gathering voter biometric information, such as facial recognition data, and putting it in a safe and immutable blockchain ledger. This ledger is virtually impossible to hack or manipulate because it is decentralized and replicated across numerous network nodes. After that, we compare the voter's identity to the information kept in the blockchain using a face recognition model based on CNN. This method enables us to protect the confidentiality and privacy of the personal data of voters while ensuring that only authorized voters can participate in the electoral process. On a real-world dataset, we put our suggested methodology to the test and show that it can accurately and securely authenticate voters. We also talk about the ethical issues that arise when implementing such a system, such as privacy, consent, and technology access. The advancement of democratic processes around the world can be aided by our suggested methodology, which has applications for creating more reliable and secure voting systems.