English(EN) Ruby-OPD: Enhancing Pronunciation Controllability and Accuracy for Flow-Matching Text-to-Speech via On-Policy Distillation

新的 TTS 框架提高了日语发音的准确性和可控性

作者 PulseAugur 编辑部 · [1 个来源] · 2026-11-05 05:00

研究人员开发了 Ruby-OPD，这是一个新颖的两阶段框架，旨在提高日语文本到语音（TTS）系统中发音的准确性和可控性。该框架解决了日语非分段特性带来的挑战，日语混合了汉字和假名，并具有广泛的同音异义词。该系统利用带有拼写扰动的 Ruby 文本表示来保持表面形式和目标发音，并在仅文本语料库上使用 On-Policy Distillation (OPD) 来提高发音准确性，而无需配对语音数据。在 Joyo-Kanji-Yomi 基准测试中，Ruby-OPD 达到了 97.13% 的准确率，展示了强大的可控性和改进的发音。 AI

影响提高了具有复杂拼写和发音的语言的 TTS 能力。

排序理由详细介绍一种新的文本到语音方法的学术论文。[lever_c_demoted from research: ic=1 ai=1.0]

在 arXiv cs.CL 阅读 →

AI 生成摘要 · Google Gemini · 来自 1 个来源。我们如何撰写摘要 →

本文如何被排名

Signal score: 26 / 100; Composite score across the factors below. Higher = stronger signal that this story matters right now.
Newsworthiness bucket: Tool; 详细介绍一种新的文本到语音方法的学术论文。[lever_c_demoted from research: ic=1 ai=1.0]
Source corroboration: Single-source cluster; Only one publisher covered this so far. Single-source stories can still rank when the publisher is high-authority, but they lack cross-source corroboration.
Topics: paper, other; Editorial topic classification. Feeds into how the story surfaces on /topic/<slug> hub pages and into the per-entity coverage mix.
AI-industry relevance: High; Clearly on-topic for AI-industry coverage.
Story freshness: Breaking (< 6h); Fresh story with cross-source coverage still developing. Ranking may shift as more sources report.

完整方法见我们的编辑标准。

报道来源 [1]

arXiv cs.CL TIER_1 English(EN) · Rui Hu, Xiaolong Lin, Zhenpeng Zhan · 2026-11-05 05:00

Ruby-OPD：通过 On-Policy Distillation 增强流匹配文本到语音的读音可控性和准确性

arXiv:2611.06403v1 Announce Type: new 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 ext…

本文如何被排名

报道来源 [1]

Ruby-OPD：通过 On-Policy Distillation 增强流匹配文本到语音的读音可控性和准确性

相关实体

相关话题