Google kicked off its I/O developer conference today by announcing an extension for Google Translate. Specifically, the service now supports 24 additional languages. This takes the overall number of supported languages to 133.
The new additions are the first to use a technique known as Zero-Shot Machine Translation, in which a machine learning model only encounters monolingual text. It learns to translate in another language without ever seeing an example of a translation.
While technically amazing, this also implies that translation accuracy is lower than in other languages, at least for the time being. Google claims that it will continue to improve the results in the future.
The newly added languages are spoken by nearly 300 million people worldwide, including 800,000 in India’s far northeast and 45 million in Central Africa. Some Indigenous languages of the Americas, as well as an English dialect, have been introduced for the first time.
The whole list is as follows: Assamese, Aymara, Bambara, Bhojpuri, Dhivehi, Dogri, Ewe, Guarani, Ilocano, Konkani, Krio, Kurdish (Sorani), Lingala, Luganda, Maithili, Meiteilon (Manipuri), Mizo, Oromo, Quechua, Sanskrit, Sepedi, Tigriny