A Method to Improve the Accuracy of Face Recognition in Application Scenarios

This paper mainly studies the method to improve the accuracy of face recognition by using the multi sample discrimination method and the automatic optimization method of the sample library through analyzing the problems in the face recognition application scenarios with continuous input data such as intelligent video analysis. In the automatic optimization process of the sample library, a method of dynamically updating the face sample library while face recognition is proposed. The essence of this method is to find the sample combination corresponding to the optimal constraint function value. The conditional constraint function contains two criteria: global center approximation and single distance approximation. In the simulation experiment, the recognition accuracy has been improved after using this method.