SPEECH RECOGNITION BY LINEAR PREDICTION

Abstract

Speech recognition is fundamentally pattern classification task. It is divided mainly into two major parts. The first part is speech signal processing and the second part is speech pattern recognition technique. The speech processing stage consists of speech starting and end point detection, pre-emphasis, frame blocking, windowing, filtering, calculating the Linear Predictive Coding (LPC) and finally constructing the codebook by vector quantization. The second part consists of pattern recognition system using Neural Network (NN). We use feed forward back propagation neural network for classification, both for speaker dependent and speaker independent system. Speech signals are recorded using an audio wave recorder in the normal room environment. The research work has been done using 50 different hybrid paired word (HPW) by 10 male and 10 female speakers. The performance of 95.755% recognition rate for speaker dependent and 70.305% recognition rate for speaker independent was established.