Research On The Technology Of ANN For SR And Application In Information Security

With the rapid development of science and technology, the information industry in the national economy and people's life plays an increasingly important role. How to effectively extract the feature of the information holder's identity is the key of guaranteeing the information security, so the personality identification based on safe access control is more and more significatively in praxis. Up to now, the technology using in speaker recognition is most based on conventional linear system theory, but the speaker recognition itself is a complex nonlinear assoeted process. Artificial neural network is a distributed parallel processing network model based on the theories of nonlinear-system, and has a strong ability of pattern classification and robust to incomplete information. It is an inevitable trend to research the speaker recognition technology based on neural network and it provides a unique approach and idea for speaker recognition technology.First, based on the research of the collection of speech, preprocessing and the common methods of feature extraction, This paper focus on Back-Propagation neural network and Self-Organizing Map neural network based on pattern recognition; then, through the experiments, the paper proposes an improved Principal Component Analysis combined with the advantages of the conventional Principal Component Analysis to optimize speaker feature and optimizes the neural network combined with many experiments and the Genetic Algorithms; finally, Experimental results show that the new approach is much more effectively to improve the rate of the speaker recognition.