A COMPARATIVE ANALYSIS ON FACE ANTI SPOOFING DETECTION APPROACHES

Abstract

With the advancements in technology, face recognition systems have become increasingly prevalent in various applications, ranging from security systems to user authentication. However, these systems are susceptible to spoofing attacks, where adversaries attempt to deceive the system by presenting manipulated or fake face images. To address this vulnerability, numerous face anti-spoofing detection approaches have been proposed. And with the rise in sophisticated spoofing attacks, it is crucial to evaluate and compare different face anti-spoofing detection approaches to identify their strengths, weaknesses, and overall performance. This thesis presents a comprehensive comparative analysis of various face anti-spoofing detection approaches, including traditional methods and deep learning-based techniques. The objective is to assess their effectiveness in detecting and differentiating genuine faces from spoofed faces, considering different types of spoofing attacks and datasets. The analysis includes evaluation metrics such as accuracy, false acceptance rate, false rejection rate, and receiver operating characteristic curves. The findings of this study provide valuable insights into the strengths and limitations of different approaches, enabling researchers and practitioners to make informed decisions when choosing face anti-spoofing techniques for real-world applications