Design And Establishment Of Romote Speaker Recognition System

Human’s biological characteristics are unique, measurable, hereditable and lifelong characteristics generally. Therefore, compared with the traditional identity recognition technology, biometrics has a great advantage.Such as speaker recognition it is more user friendly and convenient; it is simple and easy to use, it can be used for identity recognition remotely and unconsciously.What’s more, with the rapid development of information communication technology and Internet technology, the home-use computer and mobile handheld devices become popular and universal. At the same time, the remote access system has also been attracted much attention in recent years. The author thought that the above two reasons, and in the background of Chinese the elderly receiving social security deposit, developing a remote speaker recognition system is of great practical significance. Current elderly speaker recognition system has some defective. For example, applicability is not strong; anti-recording playback attack is weak; human-computer interaction is unfriendly; actual recognition rate is not high. In this paper a new type of remote speaker recognition system was designed.(1) In this paper, several key theory and technology of improving service quality of the remote speaker recognition system was discussed. Including core authentication theory of speaker recognition system and anti-recording playback theory, human-computer interaction method which is based on the voice interactive technology. In order to ensure the core authentication rate of speaker recognition system, we also adapt the strategies of collecting voice sample data(about voice volume and background noise); what’s more, a LAMP(Linux + Apache + My SQL + Python) architecture of web services platform technology was Designed,which plays an important role in improving the service quality of the remote speaker recognition system.(2) By analyzing the requirements of the elderly receiving social security, the paper presents a three-stage technological process, including identity registration, recognition data registration and speaker recognition. A system framework which is composed of information processing centers, call centers and computer client is designed, too.(3) Implemente and validate the each module of the system from the server to the client according to the system architecture: The server system and database which is based on structure design of the LAMP architecture; Setup the physical platform and the software platform of telephone man-machine terminal control center; Design the client of C / S structure system.(4) A comprehensive assessment of the system, including functional testing and performance testing, not only provide reliable experimental data for the study of remote speaker recognition system, but also ensure the usability of the system...