2008 · 20 citations · 12 references
ComputingComputer ScienceComputational LinguisticsNatural Language ProcessingPhoneticsAffective ComputingAffective State DetectionEmotion RecognitionLanguage ModelingSpeech RecognitionSuprasegmental PhoneticsEmotion DetectionEmotion Recognition From SpeechEmotional State AssessmentParalinguistic CuesParalinguistic InformationPhoneme RecognitionRobust Emotion RecognitionVoice Quality
Recognition and detection of non-lexical or paralinguistic cues from speech usually uses one general model per event (emotional state, level of interest). Commonly this model is trained independent of the phonetic structure. Given sufficient data, this approach seemingly works well enough. Yet, this paper addresses the question on which phonetic level there is the onset of emotions and level of interest. We therefore compare phoneme-, word- and sentence-level analysis for emotional sentence classification by use of a large prosodic, spectral, and voice quality feature space for SVM and MFCC for HMM/GMM. Experiments also take the necessity of ASR into account to select appropriate unit-models. In experiments on the well-known public EMO-DB database, and the SUSAS and AVIC spontaneous interest corpora, we found that the emotion recognition by sentence level analysis shows the best results. We discuss the implications of these types of analysis on the design of robust emotion and interest recognition of usable human-machine interfaces (HMI). Index Terms: emotion and interest recognition, affective speech, phoneme and word models
12
A database of German emotional speech
Felix Burkhardt, Astrid Paeschke, Manfred Rolfes et al. · 2005 · 2.2K citations
Hidden Markov model-based speech emotion recognition
Björn W. Schuller, Gerhard Rigoll, M. Lang · 2003 · 581 citations
Towards More Reality in the Recognition of Emotional Speech
Björn W. Schuller, Dino Seppi, Anton Batliner et al. · 2007 · 124 citations · Full text