A GEOSTATISTICAL MODEL FOR LINEAR PREDICTION ANALYSIS OF SPEECH

This paper presents a geostatistical model as a new approach to the li near prediction analysis oi speech. The autocorrelation method of auto regressive modeling, which is widely applied in the linear predictive coding of speech, is used as a benchmark for comparison with the prese nt algorithm. Before discussing the proposed model, we will briefly de scribe the concepts of linear prediction analysis of speech and how th is is solved by the well-known method of autocorrelation. Following is the introduction of geostatistics including the ideas of regionalized variables, semi-variograms and kriging equations. We then propose a g eostatistical model to the linear prediction modeling of speech signal s. Examples on speech data are given to illustrate the effectiveness o f the present algorithm in comparison with the autocorrelation method. Advantages offered by the proposed geostatistical algorithm over the autocorrelation method in the linear prediction analysis of speech are summarized as follows: (1) it is more effective due to the optimizati on of the kriging equations taking into account the biased condition; (2) it is more flexible by allowing different biased values for the fi tting of the signal spectrum, and therefore may provide a means for ad aptive LPC; (3) it can give a good estimate of the number of poles use d in the LPC by means of the theoretical semi-variogram. (C) 1998 Patt ern Recognition Society. Published by Elsevier Science Ltd. All rights reserved.