Biometrics-based approach for online security

• Main Author: Kaung, Myat Shwe
• Other Authors: Mohammed Yakoob Siyal
• Format: Final Year Project
• Language: English
• Published: Nanyang Technological University 2022
• Subjects: Engineering::Electrical and electronic engineering::Computer hardware, software and systems
• Online Access: https://hdl.handle.net/10356/158271
• Institution: Nanyang Technological University

Along with the new technological advancements, hackers are also evolving and continue to pose a threat to cybersecurity. In the face of rising cyberattacks, conventional methods of security such as passwords and pins are proven vulnerable and ineffective against them. Hence, biometric authentication is fast becoming the preferred way to safeguard the individuals and companies from threat actors and represent a favorable alternative to passwords. Therefore, in this project, the author will explore different biometric techniques and implement the proposed method which is face recognition. An application prototype which will combine the password and biometric feature as a two-factor authentication, is developed using MATLAB Apps Designer. Two algorithms are implemented for the purpose of face recognition: support vector machine classification and AlexNet deep transfer learning. Different sets of experimental testings are then carried out under various constraints such as illumination, face orientation, expression, occlusion, etc. Both algorithms show promising results for the task of facial recognition and it is recommended for usage throughout modern societies.