Emotion Recognition in Spontaneous Speech
2006 (English)In: Working Papers 52: Proceedings of Fonetik 2006, Lund, Sweden: Lund University, Centre for Languages & Literature, Dept. of Linguistics & Phonetics , 2006, 101-104 p.Conference paper (Other academic)
Automatic detection of emotions has been evaluated using standard Mel-frequency Cepstral Coefficients, MFCCs, and a variant, MFCC-low, that is calculated between 20 and 300 Hz in order to model pitch. Plain pitch features have been used as well. These acoustic features have all been modeled by Gaussian mixture models, GMMs, on the frame level. The method has been tested on two different corpora and languages; Swedish voice controlled telephone services and English meetings. The results indicate that using GMMs on the frame level is a feasible technique for emotion classification. The two MFCC methods have similar perform-ance, and MFCC-low outperforms the pitch features. Combining the three classifiers signifi-cantly improves performance.
Place, publisher, year, edition, pages
Lund, Sweden: Lund University, Centre for Languages & Literature, Dept. of Linguistics & Phonetics , 2006. 101-104 p.
Computer Science Language Technology (Computational Linguistics)
IdentifiersURN: urn:nbn:se:kth:diva-51948OAI: oai:DiVA.org:kth-51948DiVA: diva2:465242
Fonetik 2006. Lund, Sweden. June 7-9, 2006
tmh_import_11_12_14. QC 201112222011-12-142011-12-142011-12-22Bibliographically approved