The recently launched Huawei Mate 20 and Mate 20 Pro introduced fully neural on-device language translations via the Microsoft Translator app in partnership with Microsoft. The app comes pre-installed on Huawei Mate 20 devices allowing users to have native access to online quality level translations even when they are not connected to the internet.
According to Microsoft, the breakthrough was achieved by customizing its neural machine translation technology for the NPU (Neural Processing Unit) in the Kirin 980 SoC that powers the Mate 20 devices. This allows the smartphones to get better and faster translations compared to existing offline packs available for other devices using the power of neural network translation without an internet connection.
Microsoft uses the industry standard BLEU score to measure its translation quality. BLEU rates translation quality by comparing how close machine translations are to human ones.
Offline translation quality with Microsoft Translator on Huawei Mate 20 devices is closer to human translation.
Microsoft claims that using this methodology, depending on the language, it achieves up to a 23 percent better offline translation quality over competing best-in-class offline packs. This means that even when you’re not connected to the internet, offline translation quality using the Microsoft Translator app on Huawei Mate 20 devices is closer to human translation than offline translations have ever been.
In the internal tests, the translation of a full-page text picture was not only more accurate but also up to three times faster than with the previous versions of Microsoft Translator on other high-end devices. This new technology also saves space as the language packs for Huawei Mate 20 are 50 percent smaller than current packs.
With Huawei Mate 20 Pro coming to India next week, users in India will be able to download neural translation packs for Hindi, Telugu, Tamil, Bangla, and Urdu, with more languages to come. You’d also be able to give voice inputs in English and get translations in these Indian languages.