A Speech Recognition Method of Isolated Words Based on Modified LPC Cepstrum

Abstract

The measurement of Mel spectrum distortion is a kind of warped frequency spectrum distortion measure. Using Mel frequency scale can reflect sufficiently the nonlinear perceptive characteristic of human hearings to frequency and amplitude. It can also reflect the frequency analysis and spectrum synthesis characteristics when human hear complex sounds. Aiming at speech recognition of isolated words, an improved algorithm for normal LPC cpestrum feature is put forward in this paper. That is, LPC cpestrum (LPCC) is changed nonlinear by Mel scale according to auditory characteristic, and the LPC Mel cepstrum coefficient (LPCMCC) is used as feature parameter. The speech recognition of isolated words is carried on through using RBF neural network. The experimental results show that LPCMCC feature parameter is better than LPCC feature parameter in SNRs and recognition rate.