Research and Implementation of Weighted Face Recognition Method for Multiple Angles and Mask Occlusion

Face recognition technology has made significant progress in the context of rapid development in deep learning.However,in practical applications,accuracy and practicability of face recognition are facing new challenges due to mask occlusion and facial angle variation.Currently,face recognitions in the cases of mask occlusion and varying facial angles are processed respectively,lacking a method to comprehensively address the problem of multi-conditional face recognition affected by the two issues.Thus,considering the issues of mask occlusion and facial angle variation,a novel weighted face recognition method is proposed in this paper,where the face recognition problem is first classified according to whether wearing mask and then multiple face recognition models are employed for identification,and finally a weight determination mechanism based on face rotation angle is introduced to make better use of the models in scenarios involving face rotation and mask occlusion.The proposed method outperforms existing face recognition methods when handling the difficulties of multiple angles and mask occlusion,providing a novel solution for face recognition in the post-pandemic era.