Enhanced human facial anti-spoofing through texture analysis leveraging convolutional neural networks

Face recognition is a widely utilized authentication method in biometric systems. However, a critical challenge faced by these systems is the occurrence of false acceptance, where unauthorized individuals or attackers gain access. This type of security breach, known as a presentation attack, exposes the system to significant vulnerabilities. The focus of this research is on a specific type of presentation attack called spoofing, where attackers employ artifacts such as photographs, masks, or pre-recorded videos to replicate the appearance of legitimate users and gain fraudulent access. The primary objective of this study is to develop a neural network model capable of accurately detecting whether a user is genuine or a spoof, thereby mitigating the risk of biometric system compromise.