Speech Recognition Research Based On Quantum Neural Network Specific

Speech recognition is one of the key technology and the most ideal way for man-machine communication, which combines the theoretical achievements of many disciplines. According to the degree of dependence on the speaker, speech recognition can be divided into specific and non-specific speech recognition system. The specific speech recognition system has broad application prospects, such as industrial control, computer control, home appliance service, personal information confirm, personal mobile communication, and so on. The process of specific speaker speech recognition includes feature extraction, training and template matching.Quantum neural network is a new theory of artificial intelligence technology formed by a combination of quantum computation and ANN, which uses quantum parallel computing and quantum entanglement characteristics, to make up for the defects of the traditional ANN, promote the development of the ANN theory in essence with extremely important theoretical significance and practical significance.This paper introduces in detail the speech recognition and the process of specific speaker speech recognition. And start from the quantum neural network model,to study several major quantum neural network model,such as the multi-level transfer function quantum neural networks, the quantum transition neural networks, quantum entanglement of neural networks, quantum parallel model of self-organizing map, multi-universe quantum neural network, and analyze the structure and characteristics of the model, and focus on the multi-level transfer function quantum Neural Networks, and made a speech recognition algorithm based on the multi-level transfer function quantum Neural Network, designed based on multi-level transfer function quantum Neural network speech recognition system.In this article, the experimental results and related data are from the laboratory environment. The experimental results show that based on multi-level transfer function quantum neural network achieved good results in the recognition rate and credibility. It also reflects the multi-level transfer function quantum neural network to solve the problem of speech recognition has great advantages and potential applications.