Emotion Recognition Expressed Contextually for Romanian Language

This paper describes the results obtained by using emotional speech database contextually expressed in theater plays. The emotions analyzed in this study are happiness, sadness, anger and neutral tone. The best recognition was 73.67% and it was obtained for the hybrid feature vectors which include cepstral coefficients: PARCOR (partial correlation coef.) + LAR (log area ratio coef.) + AC (autocorrelation coef).