arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2306.00535cs.CLeess.AS
收藏

低资源文本转语音迁移学习中输入类型与发音词典使用的影响

The Effects of Input Type and Pronunciation Dictionary Usage in Transfer Learning for Low-Resource Text-to-Speech

University of Groningen(格罗宁根大学);Fryske Akademy(弗里斯兰学院);Mercator Research Centre(墨卡托研究中心);ReadSpeaker(ReadSpeaker公司)

Phat Do, Matt Coler, Jelske Dijkstra, Esther Klabbers

AI总结:

本文研究低资源 TTS 跨语言迁移的输入表示与无发音词典问题,证明发音特征优于音素标签,并提出多语言 G2P 与通用音素识别两种建典方案。

摘要:

我们比较了音素标签与发音特征作为输入,在面向低资源语言(LRL)的文本转语音(TTS)跨语言迁移学习中的表现。使用 FastSpeech 2 并以低资源语言西弗里斯兰语开展的实验表明,发音特征在可懂度和自然度上均优于音素标签。针对没有发音词典的低资源语言,我们提出两种新方法:a)在训练和合成阶段均使用大规模多语言模型进行字素到音素(G2P)转换;b)使用通用音素识别器构建临时词典。结果显示,G2P 方法的表现总体上与使用真实词典相当;音素识别方法虽然通常表现较差,但对于不太适合 G2P 方法的低资源语言仍是可行选择。在每种方法中,使用发音特征作为输入均优于使用音素标签。

英文摘要:

We compare phone labels and articulatory features as input for cross-lingual transfer learning in text-to-speech (TTS) for low-resource languages (LRLs). Experiments with FastSpeech 2 and the LRL West Frisian show that using articulatory features outperformed using phone labels in both intelligibility and naturalness. For LRLs without pronunciation dictionaries, we propose two novel approaches: a) using a massively multilingual model to convert grapheme-to-phone (G2P) in both training and synthesizing, and b) using a universal phone recognizer to create a makeshift dictionary. Results show that the G2P approach performs largely on par with using a ground-truth dictionary and the phone recognition approach, while performing generally worse, remains a viable option for LRLs less suitable for the G2P approach. Within each approach, using articulatory features as input outperforms using phone labels.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑