Research On Speech Recognition Based On Neural Networks

Speech recognition has received more and more attention recently due to the important theoretical meaning and practical value. Up to now, most speech recognition is based on conventional linear system theory, such as Hidden Markov Model (HMM) and Dynamic Time Warping (DTW). With the deep study of speech recognition, it is found that speech signal is a complex nonlinear process. If the study of speech recognition wants to break through, nonlinear system theory method must be introduced to it. Recently, with the development of nonlinear-system theories such as artificial neural networks (ANN), chaos and fractal, it is possible to apply these theories to speech recognition. Therefore, the study of this paper is based on ANN and chaos and fractal theories are introduced to process speech recognition.This paper mainly studies speech recognition based on ANN. Computing validation, performance analysis and results assessing are handled to each part of speech recognition process such as preprocessing, feature extraction and recognition algorithms. The performance of speech recognition and application characteristic of several recognition methods used in this paper is compared. The design principle of ANN and the effects of different feature parameters to speech recognition results are analyzed and discussed. The related algorithms and models are developed. The design and exploitation of software for experiments is also completed. There exists fractal in speech. Therefore, fractal dimension may be regarded as a kind of feature parameters and be combined with conventional feature parameters in speech recognition. In order to represent the feature of speech better and avoid the localization of using subsection linear method, the method for speech recognition based on mixed parameter of Mel-Frequency Cepstrum Coefficients (MFCC) and fractal dimension is proposed in this paper. Experimental results show that using mixed feature parameters achieves better effects than using MFCC feature parameters only. With above investigations, chaotic behavior in speech is studied and chaotic neural network (CNN) is applied to speech recognition. A CNN architecture is constructed based on a kind of chaotic neurons. The method for speech recognition based on CNN is proposed. Experimental results show better recognition performance and particular application advantages are...