Speech recognition using Radial Basis Function neural network

In this paper a novel approach for implementing isolated speech recognition is studied. While most of the literature on speech recognition (SR) is based on hidden Markov model (HMM), the present system is implemented by Radial Basis Function type neural network. The two phases of training and testing in a Radial Basis Function type neural network has been described. All of classifiers use Linear Predictive Cepstral Coefficients. It is found that the performance of Radial Basis Function type neural networks is superior to the other classifier Multilayer Perceptron Neural Networks. The promising results obtained through this design show that this new neural networks approach can compete with the traditional speech recognition approaches. Promising results are obtained both in the training and testing phases due to the exploitation of discriminative information with neural networks. It is found that RBF trains and tests faster than MLP.