Enregistré dans:
Détails bibliographiques
Auteurs principaux: Tee, Hitomi Jin Ling, Wang, Chaoren, Zhang, Zijie, Wu, Zhizheng
Format: Preprint
Publié: 2025
Sujets:
Accès en ligne:https://arxiv.org/abs/2510.26190
Tags: Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
_version_ 1866915586776236032
author Tee, Hitomi Jin Ling
Wang, Chaoren
Zhang, Zijie
Wu, Zhizheng
author_facet Tee, Hitomi Jin Ling
Wang, Chaoren
Zhang, Zijie
Wu, Zhizheng
contents The evaluation of intelligibility for TTS has reached a bottleneck, as existing assessments heavily rely on word-by-word accuracy metrics such as WER, which fail to capture the complexity of real-world speech or reflect human comprehension needs. To address this, we propose Spoken-Passage Multiple-Choice Question Answering, a novel subjective approach evaluating the accuracy of key information in synthesized speech, and release SP-MCQA-Eval, an 8.76-hour news-style benchmark dataset for SP-MCQA evaluation. Our experiments reveal that low WER does not necessarily guarantee high key-information accuracy, exposing a gap between traditional metrics and practical intelligibility. SP-MCQA shows that even state-of-the-art (SOTA) models still lack robust text normalization and phonetic accuracy. This work underscores the urgent need for high-level, more life-like evaluation criteria now that many systems already excel at WER yet may fall short on real-world intelligibility.
format Preprint
id arxiv_https___arxiv_org_abs_2510_26190
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle SP-MCQA: Evaluating Intelligibility of TTS Beyond the Word Level
Tee, Hitomi Jin Ling
Wang, Chaoren
Zhang, Zijie
Wu, Zhizheng
Sound
Computation and Language
Audio and Speech Processing
The evaluation of intelligibility for TTS has reached a bottleneck, as existing assessments heavily rely on word-by-word accuracy metrics such as WER, which fail to capture the complexity of real-world speech or reflect human comprehension needs. To address this, we propose Spoken-Passage Multiple-Choice Question Answering, a novel subjective approach evaluating the accuracy of key information in synthesized speech, and release SP-MCQA-Eval, an 8.76-hour news-style benchmark dataset for SP-MCQA evaluation. Our experiments reveal that low WER does not necessarily guarantee high key-information accuracy, exposing a gap between traditional metrics and practical intelligibility. SP-MCQA shows that even state-of-the-art (SOTA) models still lack robust text normalization and phonetic accuracy. This work underscores the urgent need for high-level, more life-like evaluation criteria now that many systems already excel at WER yet may fall short on real-world intelligibility.
title SP-MCQA: Evaluating Intelligibility of TTS Beyond the Word Level
topic Sound
Computation and Language
Audio and Speech Processing
url https://arxiv.org/abs/2510.26190