Generating Multilingual Voices Using Speaker Space Translation Based on Bilingual Speaker Data
This addresses the need for multilingual text-to-speech systems that maintain speaker identity, though it is incremental as it builds on existing embedding techniques.
The paper tackled the problem of enabling monolingual voices to speak a second language while preserving voice quality, using a speaker space translation method based on bilingual speaker data, and found that a simple transform achieved high naturalness, even outperforming a native voice in one listening test.
We present progress towards bilingual Text-to-Speech which is able to transform a monolingual voice to speak a second language while preserving speaker voice quality. We demonstrate that a bilingual speaker embedding space contains a separate distribution for each language and that a simple transform in speaker space generated by the speaker embedding can be used to control the degree of accent of a synthetic voice in a language. The same transform can be applied even to monolingual speakers. In our experiments speaker data from an English-Spanish (Mexican) bilingual speaker was used, and the goal was to enable English speakers to speak Spanish and Spanish speakers to speak English. We found that the simple transform was sufficient to convert a voice from one language to the other with a high degree of naturalness. In one case the transformed voice outperformed a native language voice in listening tests. Experiments further indicated that the transform preserved many of the characteristics of the original voice. The degree of accent present can be controlled and naturalness is relatively consistent across a range of accent values.