Google said the expansion brings speech recognition to over a billion new users worldwide. The list of newly supported languages includes Amernian (Armenia), Bengali (Bangladesh, India), Sundanese (Indonesia) and Urdu (Pakistan, India). Google published the full list in a blog post on the launch.
The languages cover several different areas of the world. Google said it has particularly focused on African, Asian and ancient languages. The rollout includes Swahili and Amharic, two of the most used languages in Africa. There’s also support for Georgian, a language that can be dated back to the 10th century.
Google said it’s optimistic about the impact voice recognition can have. The company cited research that found voice dictation can be three times faster than typing a message. Despite its possible value to worldwide users, it’s still largely confined to languages in developed regions such as America and Europe. Google’s trying to make its products compatible with a broader range of tongues.
Native speakers assisted in the development of the new languages by contributing sound samples. Google asked natives to read common phrases aloud, creating a baseline that its artificial intelligence systems could use to understand what the language sounds like. Machine learning was then used to work out the sounds of different words. Google said this offers improved accuracy and will be further developed as more samples are fed to the system.
READ NEXT: Facebook bans ‘cloaking’ as it cracks down on scams
“To incorporate 30 new language varieties, we worked with native speakers to collect speech samples, asking them to read common phrases,” explained Google. “This process trained our machine learning models to understand the sounds and words of the new languages and to improve their accuracy when exposed to more examples over time.”
The new languages are available now across all of Google’s voice recognition products. The dialects have been incorporated into the company’s Cloud Speech API, the service which apps like Translate and Gboard are based on. Third party developers will also be able to access the new tongues in their apps.
As well as the additional worldwide languages, Google’s today rolling out another voice feature for U.S. English users. You can now create emoji in text while using voice dictation, removing the need to return to your keyboard. If you want to add an emoji to your message, you can say its name aloud, potentially making it even quicker to express yourself in a chat.
