Speech rate conversion without extraction of accurate pitch period

Many speech rate conversion methods have been proposed. In a speech rate conversion, there are two kinds of processes: one is a contraction of a speech period and the other is an extension. Specially, the extension of speech period is useful for aged persons to listen easily without changing its composed frequencies of speech and personality. In these methods, the extraction of an accurate pitch period in a voice is needed. However, an accurate pitch period extraction is very complicated. This paper proposes a speech rate conversion without the extraction of an accurate pitch period. As a result, the constructed speech by the proposed method gives a better quality than other convention methods, according to the calculation of power weighted LPC cepstrum distance and the percent error of its output sample numbers.