Research on Liveness Detection Algorithms Based on Deep Learning

Face liveness detection is designed to prevent the face recognition system from using the attacking image as the user's real image. At present, although there are many liveness detection methods, most of the methods have poor generalization ability in practical application. Because these methods are based on datasets, and these data cannot well simulate the data distribution in the real environment. In order to improve the generalization ability of the model, in this paper, we build model using the deep neural network and the dataset collected by ourselves. The experimental results show that our model has achieved significant improvement in the realistic scenario.