Liveness control in face recognition with deep learning methods

Abstract

Today, automatic identification of individuals from biometric features is widely used in identification and authentication, security, and monitoring applications. Since facial recognition is a more user-friendly and comfortable method than other biometric methods, it has grown rapidly in recent years. However, most facial recognition systems are vulnerable to spoofing attacks. Therefore, face liveness detection (FLD) methods are of great importance. On the other hand, unlike traditional methods, deep learning techniques promise to significantly increase the accuracy of facial liveness detection systems and eliminate the difficulties of the real-world implementation of these systems. Therefore, in this paper, the application of some deep learning models to detect face liveness is reviewed and compared with each other.

Palmprint Recognition Using Pre-Trained Convolutional Neural Networks

“…Edges, shapes, and textures are just a few examples of the essential features that these models have learned to identify and extract from photos. Utilizing pre-trained CNNs can be highly beneficial for various computer vision applications, such as semantic segmentation, object recognition, and image classification [5,[14][15][16]. The following provides a summary of several pretrained models utilized in this paper.…”