Investigation of Speaker Group-Dependent Modelling for Recognition of Affective States from Speech

Siegert, Ingo; Philippou-Hübner, David; Hartmann, Knut; Böck, Ronald; Wendemuth, Andreas

doi:10.1007/s12559-014-9296-6

Cited by 23 publications

(13 citation statements)

References 77 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…As we can see, there is a huge variety in the number of selected features but also in the type of features [19]. Further, from the literature review in [20], we can state that for several of the observed corpora reasonable results were achieved by various research groups using considerably less than 6552 features. According to [7], we also decided to start with a larger feature collection.…”

Section: A Feature Extractionmentioning

confidence: 91%

Exploring dataset similarities using PCA-based feature selection

Siegert

Böck

Wendemuth

et al. 2015

2015 International Conference on Affective Computing and Intelligent Interaction (ACII)

Self Cite

View full text Add to dashboard Cite

Abstract-In emotion recognition from speech, several wellestablished corpora are used to date for the development of classification engines. The data is annotated differently, and the community in the field uses a variety of feature extraction schemes. The aim of this paper is to investigate promising features for individual corpora and then compare the results for proposing optimal features across data sets, introducing a new ranking method. Further, this enables us to present a method for automatic identification of groups of corpora with similar characteristics. This answers an urgent question in classifier development, namely whether data from different corpora is similar enough to jointly be used as training material, overcoming shortage of material in matching domains. We compare the results of this method with manual groupings of corpora. We consider the established emotional speech corpora AVIC, ABC, DES, EMO-DB, ENTERFACE, SAL, SMARTKOM, SUSAS and VAM, however our approach is general.

show abstract

Section: A Feature Extractionmentioning

confidence: 91%

Exploring dataset similarities using PCA-based feature selection

Siegert

Böck

Wendemuth

et al. 2015

2015 International Conference on Affective Computing and Intelligent Interaction (ACII)

Self Cite

View full text Add to dashboard Cite

show abstract

“…Through experiments with different datasets it was shown that the recognition performance is significantly improving when considering gender or age. In some cases a combination of both factors achieved an even further improvement [4]. Subsequently, the gender and age-group specific modelling was extended to the fusion of continuous, fragmentary, audio-visual data.…”

Section: Investigated Open Issuesmentioning

confidence: 99%

Emotional and User-Specific Acoustic Cues for Improved Analysis of Naturalistic Interactions

Siegert

2015

Künstl Intell

Self Cite

View full text Add to dashboard Cite

“…Words which are strongly emphasized by the speaker and hence are perceived as very prominent by the listener frequently indicate a correction after a misunderstanding [41]. Endowing a machine with the capabilities to use this information is the target of our and previous research [26,27]. These words are also visually prominent, i. e. when only observing and not acoutically listening to the speaker [2,5,16,29,40].…”

Section: Introductionmentioning

confidence: 99%

“…These words are also visually prominent, i. e. when only observing and not acoutically listening to the speaker [2,5,16,29,40]. Previously algorithms have been developed to detect these prominent words from the acoustic signal [26,27,33]. In [17] we presented an approach which also integrates information on the speaker's head and mouth movement to detect prominent words.…”

Section: Introductionmentioning

confidence: 99%