Automatic recognition of spontaneous emotions in speech using acoustic and lexical features
conference paper
We developed acoustic and lexical classifiers, based on a boosting algorithm, to assess the separability on arousal and valence dimensions in spontaneous emotional speech. The spontaneous emotional speech data was acquired by inviting subjects to play a first-person shooter video game. Our acoustic classifiers performed significantly better than the lexical classifiers on the arousal dimension. On the valence dimension, our lexical classifiers usually outperformed the acoustic classifiers. Finally, fusion between acoustic and lexical features on feature level did not always significantly improve classification performance. © 2008 Springer-Verlag Berlin Heidelberg.
Topics
AcousticsAudio acousticsClassifiersFlow interactionsInteractive computer systemsRobot learningSpeech analysisSpeech recognitionTechnical presentationsUser interfacesAutomatic recognitionsBoosting algorithmsClassification performancesEmotional speechesFeature levelsLexical featuresSpontaneous emotionsVideo gamesLearning systemsemotions
TNO Identifier
181301
ISSN
03029743 ; 3540858520 (ISBN); 9783540858522 (ISBN) AU -
Source title
5th International Workshop on Machine Learning for Multimodal Interaction, MLMI 2008 , 8 September 2008 through 10 September 2008, Utrecht, 74826
Pages
161 - 172
Files
To receive the publication files, please send an e-mail request to TNO Repository.