Face Spoof Detection In Artificial Neural Networks Using Deep Learning

Abstract

Face spoof detection is crucial for biometric security and the reliability of face recognition. Advancements in artificial neural networks and machine learning algorithms have produced promising results in this area. As facial recognition technology is widely used in security, authentication, and access control, there has been a rise in face spoofing incidents. This attention-enhanced face spoof detection framework was developed using a Kaggle dataset and feature extraction techniques. An ensemble of three machine learning models, a Convolutional Neural Network (CNN), an Artificial Neural Network (ANN), and a Long Short-Term Memory (LSTM), was implemented to recognize subtle differences in facial features between real and spoof faces. The deep convolutional neural network achieved a superior accuracy of 98%, while the artificial neural network achieved 63% accuracy, and the long short-term memory achieved 53% accuracy. This study improves the accuracy and reliability of fraud detection systems, benefiting biometric identity, digital criminal investigations, and security measures. The proposed technique can significantly improve the security and reliability of facial recognition systems, raising their perceived reliability and accuracy. The generated system can recognize fake faces in surveillance, access control, and identity authentication systems.