Diversity-based core-set selection for text-to-speech with linguistic and acoustic features

TL;DR

This paper introduces a diversity-based core-set selection method for text-to-speech corpora, enabling efficient data subset extraction that maintains high speech quality within practical resource constraints.

Contribution

It presents a novel diversity metric for selecting representative TTS data subsets, outperforming phoneme-balanced baselines across languages and corpus sizes.

Findings

The proposed method outperforms baseline data selection techniques.

It maintains speech quality with smaller, diverse data subsets.

Effective across multiple languages and corpus sizes.

Abstract

This paper proposes a method for extracting a lightweight subset from a text-to-speech (TTS) corpus ensuring synthetic speech quality. In recent years, methods have been proposed for constructing large-scale TTS corpora by collecting diverse data from massive sources such as audiobooks and YouTube. Although these methods have gained significant attention for enhancing the expressive capabilities of TTS systems, they often prioritize collecting vast amounts of data without considering practical constraints like storage capacity and computation time in training, which limits the available data quantity. Consequently, the need arises to efficiently collect data within these volume constraints. To address this, we propose a method for selecting the core subset~(known as \textit{core-set}) from a TTS corpus on the basis of a \textit{diversity metric}, which measures the degree to which a subset encompasses a wide range. Experimental results demonstrate that our proposed method performs significantly better than the baseline phoneme-balanced data selection across language and corpus size.