An Evaluation on Deep Learning Techniques for securing Consumer’s data Confidentiality obtained through Biometrics
NeuroQuantology · 2023
The proliferation of digitalization has resulted in a substantial increase in the production and retention of individualised data on digital mediums.The expansion of the industry necessitates heightened measures for safeguarding data privacy, in order to prevent any unauthorised breach of confidential information belonging to consumers.The utilisation of biometric authentication has surfaced as a viable resolution to this issue, owing to its capacity to furnish a superior level of security by leveraging distinct physical or behavioural attributes to authenticate users.In recent years, there has been a notable surge in interest towards deep learning techniques due to their capacity to handle vast quantities of intricate data.This renders them a potentially viable approach for augmenting biometric authentication.This paper presents an evaluation of diverse deep learning methodologies that have been suggested for safeguarding consumers' data via biometric means.The fundamentals of biometric authentication are examined, encompassing its advantages and drawbacks.Subsequently, an extensive array of scholarly literature is examined, which has explored the utilisation of deep learning methodologies in the realm of biometric authentication.The present study examines the efficacy of diverse deep learning architectures, namely Convolutional Neural Networks (CNNs), Recurrent Neural Networks (RNNs), and Generative Adversarial Networks (GANs), across multiple biometric modalities, including facial recognition, voice recognition, and fingerprint recognition.