Shulou(Shulou.com)11/24 Report--

CTOnews.com November 23, iFlytek Input 12.0 version was released recently, a new upgrade to the voice personalized dialect-free switching model, using Conformer Encoder audio coding and full-language expert system to achieve user-level personalized language recognition, can support Putonghua and more than 20 dialects switch-free speech recognition.

In addition, the off-line dialect speech recognition technology has been applied in iFlytek Input to further reduce the threshold of speech interaction. At present, iFLYTEK is also based on multilingual synthesis technology, uses massive text data to enhance small language data, expands support for speech recognition in 30 languages, and further broadens the speech input scene.

The following is the update:

[Pinyin engine] updates the engine thesaurus to improve input efficiency and accuracy

[translation engine] optimize translation languages to improve translation accuracy in China, South Korea, China and Russia

Recent updates:

[dialect pronunciation] added "dialect free switching" mode, which can understand more than 20 dialects without switching.

[foreign language pronunciation] expand recognition languages, support 30 foreign language inputs, and make multilingual communication more convenient.

[Pinyin engine] upgrade scene input engine to make multi-scene input more accurate and efficient

IFlytek Input customizes different input solutions for mobile terminals, wearable devices, smart homes, car machines and other interactive interfaces. For example, it provides convenient voice input for small screen devices such as watches, designs a "split keyboard" for folding screens, and customizes functions such as "random writing" for large screens such as tablets.

Secondly, around the intelligent life scenario, iFlytek Input puts forward a set of new input schemes based on the demand for the connection between the input service and the user. Take AI assistant as an example, iFlytek Input's AI assistant has 22 built-in AI functions, covering six input scenarios: chat, post, search, comment, evaluation, and personality expression.

For complex and diverse input scenarios, iFlytek Input 12.0 is equipped with a new AI intelligent input engine, which supports efficient, accurate and free expression in chat, music, variety shows, film and television, and life service scenarios.

In addition, for gamers, workplace talent, silver-haired people, and people with visual impairment, iFlytek Input joined the game voice keyboard, cross-screen input, elder mode, barrier-free input and other modes. Under the keyboard vision, iFlytek Input creates "super skin", which integrates the sense of technology, cool dazzle and skin aesthetics into the immersive input scene. IFlytek Input is also constantly enriching content services, such as quotation functions, from providing basic communication and work communication needs to meeting users' consumption and life needs such as online shopping, social networking, takeout and so on, to make up for the new demand for content expression.

CTOnews.com learned that iFlytek Input currently has more than 100 million daily active users, a voice penetration rate of 75%, and a 45% increase in Japanese tone usage compared with 2021.

Welcome to subscribe "Shulou Technology Information " to get latest news, interesting things and hot topics in the IT industry, and controls the hottest and latest Internet news, technology news and IT industry trends.