Introduction

 
We human beings are recognized by our unique characteristics and our traits. These traits can be behavioral or physiological in nature. At present, when we consider identification we immediately associate it directly to biometrics. Identification, in general, is done by observing and identifying unique characteristics on an individual. These characteristics can be features such as our facial expressions, iris patterns, voice and even our fingerprints. In the past, we have theorized many methodologies for identifying individuals based on their unique features but were left as theories due to our lack of computation power and speed. But in this day and age, we have converted this theory and made it available for practical day to day use. \cite{thakkar}
Previously, Identification and validation for biometrics were done by information-based systems which make use of passwords or cards for authentication. Since these methods have become unreliable and less secure the information stored can be easily be hacked or lost. Therefore, there has been a rise in using more secure and complex unique identifiers for user authentication such as fingerprints for biometrics, facial recognition and so on. Because of the mentioned risks people have adopted fingerprint sensors to be used for biometric authentication for many online payment platforms and for user authentication. \cite{thakkar}
These biometric devices have small areas to scan fingerprints as the scanning area is relatively small in size. Due to its small size, compact fingerprint scanners with minimal real estate could be used alongside applications. Our fingerprints have different patterns that make each individual fingerprint unique. The patterns are made by a combination of several dermal ridges. These ridges are molded during our time in the womb, where several factors such as friction, maternal conditions, etc affect their final shape and structure. These patterns develop all over the human body including the palms, soles and even toes. Since so many conditions and factors play an important role in determining the final ridge pattern, we consider fingerprint patterns to be unique to each and every individual. In our many years of research, there has yet to be an incident where two fingerprints from different individuals have been found to be a match.
Biometrics for fingerprint authentication has become a norm in our day to day society that its vulnerability is never challenged. But due to the progress of our society in terms of technological advancements, there has been an increased risk of malicious attempts to bypass and take advantage of biometrics in general by the use of fake fingerprints. Many systems today use algorithms that match records from the database with inputs provided by the scanner for user authentication. Since these methods in specific applications are not modified and updated regularly at a pace that equals or exceeds as compared to the progress made by malicious individuals, It leaves biometrics at an increased risk and makes them susceptible to cyber attacks.
Nowadays there are many fingerprint-based systems that authenticate at high speeds that are because of great new and improved algorithms. When there is a sample data provided with closely similar features it mostly leads to false authentication. We heavily rely on man-made algorithms to provide us with great results.  In this research, we will be mainly focusing mainly on fingerprints for biometrics and how we can further improve its security and show the performance of using enhanced verification platforms alongside deep learning structures over traditional systems.
We will also detail the methodology, description of the database used in this work, architecture of the adopted networks,  parameters for training, and the evaluation protocol. Finally, the results will be presented and discussed, and we shall arrive at a conclusion.