Introducing new multilingual voices supporting 41 languages and accents with auto language detection
Published Aug 08 2023 02:14 AM 13.1K Views
Microsoft

By Jingzhou Yang, Zheng Niu, Qinying Liao, Garfield He, Jun-Wei Gan, Lei He, Binggong Ding, and Sheng Zhao

 

In today's interconnected world, language barriers can be a significant challenge when it comes to effective communication. However, with the rapid advancements in artificial intelligence and natural language processing, we are witnessing groundbreaking solutions that are revolutionizing the way we interact with technology. Azure AI Text to Speech, a powerful cloud-based service offered by Microsoft, is at the forefront of this transformation. In specific, the JennyMultilingual voice we introduced earlier has enabled customers to reach a more global audience while keeping a consistent persona across locales.

 

Initially, the JennyMultilingual voice offered support for 14 languages, facilitating communication across diverse linguistic landscapes. Now, recognizing the importance of catering to a wider audience and responding to the requests and feedback from our partners, we have taken a significant leap forward by extending the multilingual voice capability to a staggering set of 41 languages and accents.

 

In addition to the existing female voice (JennyMultilingual), we now offer a male voice (RyanMultilingual) as part of its multilingual portfolio. This addition not only provides users with more options but also enables them to create inclusive and diverse user experiences. By introducing a male voice, Azure Text to Speech ensures that the technology remains adaptable to different contexts and user preferences, empowering developers to design solutions that resonate with a broader range of end-users.

 

Moreover, both new voices come with the auto language prediction capability for the input text, eliminating the need for manual tagging as the voice can automatically recognize the input languages and adjust the speech output accordingly.

 

These new updates will be in public preview in three regions: East US, West Europe, and Southeast Asia.

 

Release details

 

New voice

With this release, we Introduce a new male multilingual voice, RyanMultilingualNeural, with a capability to speak out 41 languages/accents.

 

New languages/accents support

For both JennyMultilingual and RyanMultilingual, the default language will be en-US. Besides the 14 languages/accents which JennyMultilingual is already good at, we now extended the multilingual support to 41 languages and accents.

 

Check out the demo below to hear how Jenny and Ryan speak different languages fluently with a smooth switch.

 

The full supported language list can be found at here.

 

New runtime feature: auto language detection

By analyzing each sentence's language, the multilingual TTS system can dynamically adapt its pronunciation, intonation, and phonetics to match the specific linguistic nuances of different languages. This breakthrough allows users to effortlessly switch between languages in the text input, creating a more immersive and authentic multilingual experience. Where needed, the locale tag can still be applied using SSML, as described here.

 

Get started

Azure Neural TTS has introduced JennyMultilingual and RyanMultilingual with 41 languages and accents supported in three regions: East US, West Europe, and Southeast Asia for preview. To explore these new capabilities, simply sign up for the Speech service on Azure and access the Speech Studio Voice Gallery.

 

Microsoft provides a wide range of neural voices, offering over 400 options in more than 140 languages and locales. These text-to-speech voices enable you to quickly integrate read-aloud functionality into your applications for enhanced accessibility. They also empower chatbots to deliver more engaging and natural conversations to your users. Additionally, with the Custom Neural Voice capability, you can easily create a unique brand voice tailored specifically for your business. Using the recently announced cross-lingual capability for Custom Neural Voice, you can also enable your voice to speak dozens of different languages.

 

For more information:

 

6 Comments
Version history
Last update:
‎Sep 20 2023 03:49 AM
Updated by: