Semi-supervised methods for exploring the acoustics of simple productive feedback

Semi-supervised methods for exploring the acoustics of simple productive feedback

► Semi-automatic annotation using a dual-channel decoder gave a recall rate of 95%. ► The generalised functional distance correlates with prosodic distance: r=0.46. ► The specified functions correlates to the generalised functional distance: r=0.54. ► Functions can be predicted from prosodic cues us...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:Speech communication 2013-03, Vol.55 (3), p.451-469
Hauptverfasser: Neiberg, Daniel, Salvi, Giampiero, Gustafson, Joakim
Format: Artikel
Sprache:eng
Schlagworte:
Acoustics
Affective annotation
Correlation analysis
Feedback
feedback modeling
Feedback modelling
Grounding
grounding 2000 MSC
Human
Mathematical analysis
Mathematical models
Perception
Similarity
Social signal processing
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:► Semi-automatic annotation using a dual-channel decoder gave a recall rate of 95%. ► The generalised functional distance correlates with prosodic distance: r=0.46. ► The specified functions correlates to the generalised functional distance: r=0.54. ► Functions can be predicted from prosodic cues using regression 0.35
ISSN:0167-6393
1872-7182
1872-7182
DOI:10.1016/j.specom.2012.12.007