Vocal Emotion Recognition Research Articles

Ensemble methods for spoken emotion recognition in call-centres

Machine-based emotional intelligence is a requirement for more natural interaction between humans and computer interfaces and a basic level of accurate emotion perception is needed for computer systems to respond adequately to human emotion. Humans convey emotional information both intentionally and unintentionally via speech patterns. These vocal patterns are perceived and understood by listeners during conversation. This research aims to improve the automatic perception of vocal emotion in two ways. First, we compare two emotional speech data sources: natural, spontaneous emotional speech and acted or portrayed emotional speech. This comparison demonstrates the advantages and disadvantages of both acquisition methods and how these methods affect the end application of vocal emotion recognition. Second, we look at two classification methods which have not been applied in this field: stacked generalisation and unweighted vote. We show how these techniques can yield an improvement over traditional classification methods.

Disgust in pre-clinical Huntington's disease: A longitudinal study

Emotion recognition from both face and voice and experience of emotions were investigated in a group of non-symptomatic people at risk of carrying the Huntington's disease gene who presented for genetic testing. Based on the results of the DNA test, a group of people carrying the Huntington's disease gene (HD+), and a group of non-carriers (HD−) were formed. Since we were especially interested in the time course of possible deficits in emotion recognition, all people at risk were reassessed 6 and 12 months after the initial assessment. Recognising facial expressions of disgust was significantly impaired on all three assessments in the HD+ group, while recognition of vocal emotions and the experience of emotions were largely unaffected, confirming that deficits in recognition of facial expressions of disgust are an early correlate of carrying the gene for Huntington's disease. The inclusion of a healthy control group ( n = 37) further allowed an estimate of the genetic and environmental contribution to deficits in facial emotion recognition.

Face and emotion processing in frontal variant frontotemporal dementia

Lavenu et al. [Alzheimer Dis. Assoc. Disorder 5 (1999) 96] have shown that patients with frontotemporal dementia (FTD) show impaired recognition of facial expressions. It is not clear, however, whether these deficits arise from an impairment affecting face processing generally, emotion processing generally, or facial expression recognition alone. We address this issue by testing six patients with frontal variant frontotemporal dementia (fvFTD) on a series of face perception tasks (including facial identity and facial expression recognition), and a test of vocal emotion recognition. In general, the fvFTD participants showed impaired recognition of facial expressions in the context of preserved recognition of facial identity. In addition, however, deficits were also observed for the vocal emotion recognition task. These results are consistent with the idea that fvFTD affects the recognition of emotional signals from multiple modalities rather than facial expression processing alone. It is plausible that the emotion recognition impairments observed contribute to the abnormal social behaviour that is characteristic of this condition.

L'expression vocale des émotions : approche interculturelle et développementale

Summary : Vocal expression of emotions : A crosscultural and developmental approach. The vocal expression of emotions, an unexplored field of research, is studied here in a double approach : crosscultural and developmental. The vocal emotion recognition test (Scherer and Wallbott, 1989) mode up of 30 items depicting the emotions fear, anger, joy, sadness as well as utterances considered to be neutral, has been applied to 50 french adults comparatively to a group of german adults on which the test was first validated. This procedure was then adapted to a population of children. Three groups of children and teenagers aged 7, 11 and 15 years were tested. For the adults, the results do not show any major difference between the two cultural groups studied. A developmental evolution appears between 7 and 15 years in the recognition of the vocal stimuli. In the 15 years group, the results are quite similar to those obtained by adults. Key-words : vocal expression of emotion, non verbal communication, test.