Biometric Network Security Enhancements through Deep Learning Techniques
Renu Yadav, Kumar Sanjeev, Rahul Yogi · 2023
In Biometrics is a relatively new authentication mechanism that has garnered some acceptance in the rapidly evolving field of network security. People are increasingly interested in biometric-based security solutions since traditional systems, which usually rely on tokens or passwords, are insecure. Although biometrics have a lot of potential, several issues must be addressed before they can be used successfully, such as the data's inherent unpredictability and the technology's vulnerability to hostile attacks. With our proposed technique, we intend to overcome these issues and improve biometric network security. Our methodology is built on the use of deep learning techniques, notably for feature extraction. Deep Feature Extraction (DFE) is a technique developed by our group that employs a convolutional neural network (CNN). Unlike in the past, when human feature engineering was required, CNNs are now capable of automatically and adaptively learning the spatial hierarchies of features in biometric data. Important characteristics needed for effective identification or verification are deleted. One of the key disadvantages of deep learning models is their vulnerability to adversarial assaults, which is especially problematic for applications that manage sensitive data. It ensures enhanced accuracy for the system through the right integration of information from several sources, resulting in fewer false positives and false negatives. The proposed technique provides a comprehensive answer to the problems that biometric authentication systems face. By combining numerous algorithm combinations and cutting-edge deep learning methodologies, we want to set a new standard for biometric network security.