Speech emotion recognition using spectral features

In order to improve the performance of the speech emotion recognition system and reduce the computing complexity, a speech emo- tion recognition based on optimized coefficients number for spectral fea- tures is proposed. Experimental studies are performed over the Berlin emotional Database, using support vector machine (SVM) classifier and five spectral features MFCC, LPC, LPCC, PLP, and PLP-RASTA. The experiment result shows that the speech emotion recognition based on coefficients number can improve the performance of the emotion recog- nition system effectively. abstract environment