arXiv cs.CL· Rui Hu, Xiaolong Lin, Zhenpeng Zhan·· 11 小时前AI 评分20
Ruby-OPD：通过同策略蒸馏提升流匹配 TTS 的发音可控性与准确率
Ruby-OPD: Enhancing Pronunciation Controllability and Accuracy for Flow-Matching Text-to-Speech via On-Policy Distillation
AI 导读
研究提出两阶段流匹配 TTS 框架 Ruby-OPD，先用带拼写扰动的 Ruby 文本表示同时保留表层字形与目标发音，再在大规模纯文本语料上做同策略蒸馏（OPD），无需配对语音即可提升原始文本的发音准确率。
正文 · 原文
Abstract:Achieving both high pronunciation accuracy and fine-grained controllability remains a central challenge in Japanese text-to-speech (TTS), as Japanese is an unsegmented language mixing logographic Kanji and phonographic Kana with extensive polyphony. Traditional cascaded systems convert text into phoneme sequences via a front-end, ensuring pronunciation control but discarding lexical semantics and prosodic cues in surface orthography. Conversely, recent zero-shot TTS models directly consume raw orthographic text to improve naturalness, yet suffer from polyphone errors and lack explicit pronunciation control. To bridge this gap, we propose a two-stage framework for flow-matching TTS. First, we investigate Ruby text representations with orthography perturbation to jointly preserve surface forms and target pronunciations for reliable control. Second, we introduce On-Policy Distillation (OPD) on large-scale text-only corpora to enhance pronunciation accuracy on raw orthographic text without paired speech. On the Joyo-Kanji-Yomi benchmark, our model achieves 97.13% Accuracy, 2.02% Target PER, 0.70% Sentence PER, and 4.21% Sentence CER under raw text while retaining strong controllability.
| Comments: | 5 pages, 3 tables, work in progress |
| Subjects: | Computation and Language (cs.CL) |
| Cite as: | arXiv:2611.06403 [cs.CL] |
| (or arXiv:2611.06403v1 [cs.CL] for this version) | |
| https://doi.org/10.48550/arXiv.2611.06403 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Rui Hu [view email]
[v1]
Wed, 4 Nov 2026 12:05:34 UTC (35 KB)
来源：arXiv cs.CL · arxiv.org