What you need to know
- The latest version of the Google Translate app includes several major improvements to the instant camera translation feature.
- The feature now supports more than 60 new languages, including Arabic, Greek, Hindi, Latin, Malay, Persian, Thai, and Vietnamese.
- Thanks to Neural Machine Translation (NMT), Google claims it has been successful in reducing the number of translation errors by up to 85 percent.
Google has begun rolling out a new update for its Translate app, adding several major improvements to the camera translation feature. Perhaps the biggest improvement is the addition of support for 60 new languages, which means it is now possible for users to translate a total of 88 languages simply by pointing their camera at the text they want translated. Some of the new languages that are now supported include Arabic, Bengali, Estonian, Greek, Hindi, Javanese, Latin, Malay, Persian, Thai, and Vietnamese.
In addition to supporting newer languages, the camera translation feature can now translate text in any of 88 supported languages to any of the 100+ languages supported by Google Translate. Previously, the feature only supported translating text to English. Accuracy is another area where the camera translation feature has received a major boost with this new update. Google says translation errors have come down by up to 85 percent, all thanks to Neural Machine Translation (NMT). Until now, Google had been using the AI translation technology only in Google Lens and the web version of Google Translate.
Using the camera translation feature is also simpler now, thanks to a revamped interface and the ability to detect languages automatically. As long as you have the source set to "Detect language", the instant camera translation feature will automatically identify the language that needs to be translated. As noted by The Verge, the update will only be made available to around one percent of users today. It may take at least a few weeks for Google to make the improvements available to a larger number of users.