Local Style Tokens: Fine-Grained Prosodic Representations For TTS Expressive Control - Archive ouverte HAL Accéder directement au contenu
Communication Dans Un Congrès Année : 2023

Local Style Tokens: Fine-Grained Prosodic Representations For TTS Expressive Control

Résumé

Neural Text-To-Speech (TTS) models achieve great performances regarding naturalness, but modeling expressivity remains an ongoing challenge. Some success was found through implicit approaches like Global Style Tokens (GST), but these methods model speech style at utterance-level. In this paper, we propose to add an auxiliary module called Local Style Tokens (LST) in the encoder-decoder pipeline to model local variations in prosody. This module can implement various scales of representations; we chose Word-level and Phoneme-level prosodic representations to assess the capabilities of the proposed module to better model sub-utterance style variations. Objective evaluation of the synthetic speech shows that LST modules better capture prosodic variations on 12 common styles compared to a GST baseline. These results were validated by participants during listening tests.
Fichier principal
Vignette du fichier
lenglet23_ssw.pdf (382.47 Ko) Télécharger le fichier
Origine : Fichiers éditeurs autorisés sur une archive ouverte

Dates et versions

hal-04257713 , version 1 (25-10-2023)

Identifiants

Citer

Martin Lenglet, Olivier Perrotin, Gérard Bailly. Local Style Tokens: Fine-Grained Prosodic Representations For TTS Expressive Control. SSW 2023 - 12th ISCA Speech Synthesis Workshop (SSW2023), Gérard Bailly; Olivier Perrotin; Thomas Hueber; Damien Lolive; Nicolas Obin, Aug 2023, Grenoble, France. pp.120-126, ⟨10.21437/SSW.2023-19⟩. ⟨hal-04257713⟩
63 Consultations
83 Téléchargements

Altmetric

Partager

Gmail Facebook X LinkedIn More