Emotion Recognition from Speech Signals Using Elicited Data and Fuzzy KDA

Autor: Cheng Wei Hang, Yong Qiang Bao, Li Zhao
Rok vydání: 2013
Předmět:
Zdroj: Applied Mechanics and Materials. :1385-1388
ISSN: 1662-7482
DOI: 10.4028/www.scientific.net/amm.385-386.1385
Popis: In this paper we introduced the application of Fuzzy KDA in speech emotion recognition using elicited data. The emotional data induced in a psychology experiment. The acted data is not suitable for developing real world applications and by using more naturalistic data we may build more reliable system. The emotional feature set is then constructed for modeling and recognition. A total of 372 low level acoustic features are used and kernel discriminant analysis is used for emotion recognition. The experimental results show a promising recognition rate.
Databáze: OpenAIRE