A Convenient Authentication Algorithm Based on Multimodal Evidence

Aiming at the requirements of high reliability and convenience of identity authentication in public scene services, a multimodal identity authentication algorithm based on face, voiceprint and gesture is studied. Firstly, a multimodal identity authentication framework based on cloud model and evidence theory is established; secondly, a variety of biometric input modes are integrated, including face, voiceprint, and gesture, to provide users with more flexible identity authentication choices; finally, fusion authentication of travelers' identities is carried out through decision-level algorithms, and an evidence fusion algorithm based on the combination of cloud model and improved evidence theory is investigated to combine the recognition results of these three modal networks Input into the multi-data fusion algorithm, through the evidence fusion algorithm to achieve the final decision-making fusion, so as to further improve the reliability and rapidity of information acquisition, and realize the traveler's identity authentication. After experimental evaluation multimodal identity authentication method is significantly better than the unimodal method, and the experimental results verify the accuracy of the multimodal fusion authentication algorithm.