Feature Extraction of Spoken Dialogs for Emotion Detection
Z. Ciota · 2006
In this paper the possibilities of modern speech processing has been described. One of the most important tasks is a proper definition of feature parameters. The speaking fundamental frequencies and time-energy distribution parameters permit to create emotion recognition system. The proposed method permits to compare specific features of the current speaker voice with the computer base containing stored templates for different human emotions. The possibilities of different decision algorithm realization for emotion classification have been also discussed