Confidence-based Filtering for Speech Dataset Curation with Generative Speech Enhancement Using Discrete Tokens

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Yamauchi, Kazuki, Murata, Masato, Seki, Shogo
Format: Preprint
Published: 2026
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866914263019290624
author Yamauchi, Kazuki
Murata, Masato
Seki, Shogo
author_facet Yamauchi, Kazuki
Murata, Masato
Seki, Shogo
contents Generative speech enhancement (GSE) models show great promise in producing high-quality clean speech from noisy inputs, enabling applications such as curating noisy text-to-speech (TTS) datasets into high-quality ones. However, GSE models are prone to hallucination errors, such as phoneme omissions and speaker inconsistency, which conventional error filtering based on non-intrusive speech quality metrics often fails to detect. To address this issue, we propose a non-intrusive method for filtering hallucination errors from discrete token-based GSE models. Our method leverages the log-probabilities of generated tokens as confidence scores to detect potential errors. Experimental results show that the confidence scores strongly correlate with a suite of intrusive SE metrics, and that our method effectively identifies hallucination errors missed by conventional filtering methods. Furthermore, we demonstrate the practical utility of our method: curating an in-the-wild TTS dataset with our confidence-based filtering improves the performance of subsequently trained TTS models.
format Preprint
id arxiv_https___arxiv_org_abs_2601_12254
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Confidence-based Filtering for Speech Dataset Curation with Generative Speech Enhancement Using Discrete Tokens
Yamauchi, Kazuki
Murata, Masato
Seki, Shogo
Sound
Audio and Speech Processing
Generative speech enhancement (GSE) models show great promise in producing high-quality clean speech from noisy inputs, enabling applications such as curating noisy text-to-speech (TTS) datasets into high-quality ones. However, GSE models are prone to hallucination errors, such as phoneme omissions and speaker inconsistency, which conventional error filtering based on non-intrusive speech quality metrics often fails to detect. To address this issue, we propose a non-intrusive method for filtering hallucination errors from discrete token-based GSE models. Our method leverages the log-probabilities of generated tokens as confidence scores to detect potential errors. Experimental results show that the confidence scores strongly correlate with a suite of intrusive SE metrics, and that our method effectively identifies hallucination errors missed by conventional filtering methods. Furthermore, we demonstrate the practical utility of our method: curating an in-the-wild TTS dataset with our confidence-based filtering improves the performance of subsequently trained TTS models.
title Confidence-based Filtering for Speech Dataset Curation with Generative Speech Enhancement Using Discrete Tokens
topic Sound
Audio and Speech Processing
url https://arxiv.org/abs/2601.12254