Speech Emotion Recognition Cross Language Families: Mandarin vs. Western Languages

An investigation on classification of emotional speech cross different language families is proposed in this paper. Datasets on three languages, CDESD in Mandarin, Emo-DB in German, and DES in Danish are analyzed. With 2-D classifications on arousal-appraisal space, better recognition performances are observed in arousal dimension than in appraisal dimension. The classification rates in cross language family test between CDESD and Emo-DB or DES are far higher than chance level, shows that there exist universal mechanisms in human voice emotion independent on languages. Results in test within the same language family between Emo-DB and DES are even better than in cross language family test with CDESD in Mandarin, shows the language and culture also influence the way of expression in speech. The best classification rate in the cross language family test is achieved on male speech samples as 71.62%, when CDESD dataset is used as training set and Emo-DB as testing set.