Estimation of Emotion from Speech through Analysis of Fundamental Frequency Derivative

In this paper, we conducted an analysis of speaker emotions through speech evaluation. The Introduction section introduces an algorithm designed to assess emotional speech, which is grounded in the examination of the fundamental frequency path (F<inf xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">0</inf>). This assessment employs a decision line derived from the first derivative of the frequency, utilizing a positive and negative frequency increment. The CTD algorithm that establishes criteria for detecting emotions in plane $\left( {\begin{array}{ll} {{F_0},}&{F_{0, + - }^{(1)}} \end{array}} \right)$ is described. The decision criterion distinguishing between Normal and Boredom emotions was defined during the preliminary training phase based on the testing emotions from Base. The next section of the paper presents an experiment focusing on the statistical evaluation of accuracy concerning speaker emotion classification. To evaluation the efficacy of emotional estimations, a confusion matrix was utilized. The results from these experiments are presented both in tabular and graphical formats.