A New Statistic Feature of the Short-Time Amplitude Spectrum Values for Human's Unvoiced Pronunciation

TL;DR

This paper introduces a novel statistical feature of the short-time amplitude spectrum for unvoiced speech, revealing relationships between amplitude averages and standard deviations across frequencies, supported by a new association model.

Contribution

The paper discovers a new statistical feature of short-time amplitude spectra and proposes a model for amplitude distribution associations across frequencies in unvoiced speech.

Findings

Identified a relationship between amplitude average and standard deviation for each frequency.

Proposed a new model representing amplitude distribution associations across frequencies.

Provided mathematical proof supporting the new statistical feature and model.

Abstract

In this paper, a new statistic feature of the discrete short-time amplitude spectrum is discovered by experiments for the signals of unvoiced pronunciation. For the random-varying short-time spectrum, this feature reveals the relationship between the amplitude's average and its standard for every frequency component. On the other hand, the association between the amplitude distributions for different frequency components is also studied. A new model representing such association is inspired by the normalized histogram of amplitude. By mathematical analysis, the new statistic feature discovered is proved to be necessary evidence which supports the proposed model, and also can be direct evidence for the widely used hypothesis of "identical distribution of amplitude for all frequencies".