China unveils AI-powered Tibetan input method — ECNS
In Qinghai Province in northwestern China, Qinghai Normal University has unveiled a multilingual and multimodal Tibetan input method using artificial intelligence. The system combines text and voice input, image recognition, and AI functions, ECNS reports.
Operation on smartphones and computers
The development was created by the State Key Laboratory of Tibetan Intelligence. The input method is available for smartphones and computers. According to ECNS, it is China’s first multilingual multimodal system of this type for the Tibetan language.
The voice input function supports spoken and written forms of the Ü-Tsang, Amdo, and Kham dialects. It can also work with mixed Chinese-Tibetan-English speech.
More current news is available on the UA.News Telegram channel Telegram.
Text recognition and translation
The image recognition function can identify mixed Chinese-Tibetan-English text in photographs and screenshots. The system also offers automatic correction, text enhancement, and two-way translation between Tibetan and Chinese.
The system was equipped with terminology databases for 23 fields, including education, law, and literature. It can learn the words and expressions most frequently used by a user and also supports processing directly on the device and through cloud services.
