Temporal Features of Productive and Reproductive Speech of Speaker

Abstract

The article presents the results of a study aimed at determining the degree of influence of speech rate on the listener's differentiation of the speaker's speech type based on the productivity/reproductivity feature. This topic is especially relevant in connection with the development of modern technologies for perceiving human speech and generating artificial speech. We are at the point in the development of these technologies when machine-generated speech is still easily recognized by the listener as artificial, although the similarity to productive human speech remains great. This allows us to put forward a hypothesis that there are speech parameters that allow the listener to perceive it as productive, reproductive, or artificial. In this article, we present the results of the first part of the study, which touched upon the search for differentiating parameters for productive and reproductive speech. The study was conducted on the materials of audio recordings of productive and reproductive speech of two male speakers with different behavioral characteristics, including those expressed in speech. The dependence of the speech productivity feature on the speech rate was established. A concomitant dependence of the speech rate on the emotional state and intention of the speaker was revealed. A description of the results obtained in determining the differences between the rate and speed of the speaker's speech is given.