A NEURAL-NETWORK-BASED APPROACH TO SPEECH SIGNAL PREDICTION
Rui J. P. de Figueiredo · 2010
In this paper a new approach to speech signal prediction is presented. The approach deploys an EKF (Extended Kalman Filter)-based learning algorithm for this application. Simulation results show that the proposed neural network approach leads to better performance than the well-known linear predictive coding (LPC) approach which uses the Levinson-Durbin algorithm for predictor design.