Analysis of Emotional State of a Person and Its Effect on Speech Features Using PRAAT Software

Author

Kumbhakarn, Mansi ; Sathe-Pathak, Bageshree

Author_Institution

Electron. & Telecommun. Eng., Cummins Coll. of Eng. for women, Pune, India

fYear

2015

fDate

26-27 Feb. 2015

Firstpage

763

Lastpage

767

Abstract

Emotion Recognition and classification is emerging as an inevitable sub-system in human-machine interaction. In this paper, various features like pitch, formant frequencies, intensity, standard deviation, energy and duration have been extracted for emotion recognition and classification. Praat, which is open source scientific software for speech analysis, has been used to extract the features. Epoch count is also used, which is calculated using Zero Frequency Filtering method in MATLAB R2013a. Here, four different emotions namely Neutral, Angry, Happy and Sad are used for analysis. Effect of the above mentioned features on these four emotions have been analyzed and compared. Finally, significance of every feature for different emotions has been studied and analyzed.

Keywords

emotion recognition; feature extraction; filtering theory; human computer interaction; image classification; public domain software; speech recognition; Matlab R2013a; PRAAT software; angry emotions; emotion classification; emotion recognition; epoch count; happy emotions; human-machine interaction; neutral emotions; open source scientific software; person emotional state analysis; sad emotions; speech feature extraction analysis; zero frequency filtering method; Emotion recognition; Feature extraction; Software; Speech; Speech recognition; Standards; Vibrations; Emotion recognition and Classification; Emotions; Epochs; PRAAT software; Speech Features;

fLanguage

English

Publisher

ieee

Conference_Titel

Computing Communication Control and Automation (ICCUBEA), 2015 International Conference on

Conference_Location

Pune

Type

conf

DOI

10.1109/ICCUBEA.2015.152

Filename

7155950