Quality prediction of synthesized speech based on perceptual quality dimensions

•The feasibility of non-intrusive quality prediction for synthetic speech, which only requires acoustical measurements, is shown.•Diagnostic structuring of the perceptual quality space of TTS sound through perceptual quality dimensions.•A new framework for non-intrusive quality assessment, denoted a...

Ausführliche Beschreibung

Bibliographische Detailangaben

Veröffentlicht in:	Speech communication 2015-02, Vol.66 (Feb), p.17-35
Hauptverfasser:	Norrenbrock, Christoph R., Hinterleitner, Florian, Heute, Ulrich, Möller, Sebastian
Format:	Artikel
Sprache:	eng
Schlagworte:	Dimensional measurements Feasibility Linguistics Mathematical models Model reference Non-intrusive quality assessment Nonlinearity Perceptual regularization Speech Speech recognition Speech-quality prediction Synthetic speech perception Text-to-speech Training
Online-Zugang:	Volltext
Tags:	Tag hinzufügen Keine Tags, Fügen Sie den ersten Tag hinzu!