Active Learning for Text-to-Speech Synthesis with Informative Sample Collection

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Seki, Kentaro, Takamichi, Shinnosuke, Saeki, Takaaki, Saruwatari, Hiroshi
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866916838591430656
author Seki, Kentaro
Takamichi, Shinnosuke
Saeki, Takaaki
Saruwatari, Hiroshi
author_facet Seki, Kentaro
Takamichi, Shinnosuke
Saeki, Takaaki
Saruwatari, Hiroshi
contents The construction of high-quality datasets is a cornerstone of modern text-to-speech (TTS) systems. However, the increasing scale of available data poses significant challenges, including storage constraints. To address these issues, we propose a TTS corpus construction method based on active learning. Unlike traditional feed-forward and model-agnostic corpus construction approaches, our method iteratively alternates between data collection and model training, thereby focusing on acquiring data that is more informative for model improvement. This approach enables the construction of a data-efficient corpus. Experimental results demonstrate that the corpus constructed using our method enables higher-quality speech synthesis than corpora of the same size.
format Preprint
id arxiv_https___arxiv_org_abs_2507_08319
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Active Learning for Text-to-Speech Synthesis with Informative Sample Collection
Seki, Kentaro
Takamichi, Shinnosuke
Saeki, Takaaki
Saruwatari, Hiroshi
Sound
Audio and Speech Processing
The construction of high-quality datasets is a cornerstone of modern text-to-speech (TTS) systems. However, the increasing scale of available data poses significant challenges, including storage constraints. To address these issues, we propose a TTS corpus construction method based on active learning. Unlike traditional feed-forward and model-agnostic corpus construction approaches, our method iteratively alternates between data collection and model training, thereby focusing on acquiring data that is more informative for model improvement. This approach enables the construction of a data-efficient corpus. Experimental results demonstrate that the corpus constructed using our method enables higher-quality speech synthesis than corpora of the same size.
title Active Learning for Text-to-Speech Synthesis with Informative Sample Collection
topic Sound
Audio and Speech Processing
url https://arxiv.org/abs/2507.08319