Analysis On The Effect Of Emotion Of Intonation Based Features With Reference To Assamese Emotional Speech Signals

Abstract

Speech is a communicational source of exchanging information between human begins. Emotional speech is another precious source through which human begins expressed their internal feelings. Emotion is promoted by some particular situations, and the emotion recognition from human speech  is  considered as a most vital  subject in the field  of human-computer interfaces (HCIs) to identify with  people. Analyzing on Pitch and Formant of different gender speakers in different emotions such as Angry and normal and two different software,  in Assamese language for observing the variation between them by applying Two-way ANOVA testing is presented in this paper. The main objective of current work is to detect the effect of emotions, sex and Software’s on extracted features from recorded Assamese emotional speech of different gender in selected age group. Results of analysis of variance testing varies that emotion and sex both has significantly effect on both extracted features in recorded voices by distinct software.