The leader of the Chinese phonetic industry, Hkust has also launched a new generation of "voice cloud" platforms

A few days ago, the Chinese voice industry leader of Hkust Flying also launched a new generation of "voice cloud" platform release and Speech Developers Conference (please click: http://2012.voicecloud.cn). This new generation of "voice cloud" platform will be a natural language understanding, personalized speech recognition, personalized speech synthesis, voice recognition, oral evaluation capabilities, such as the new generation of "voice cloud" platform. Let's take a look at these five traits today.

Natural language Understanding

Natural Language, the language that people speak in the most natural way. Natural language understanding, as a new feature in the next generation of "voice cloud" platform, in the existing mobile Internet case, we can only think of the Apple Siri, do not rigidly adhere to the voice of intelligent speech control. As a pioneer, Siri successfully detonated the mobile internet voice of this bomb, set off the climax of voice applications. If the new generation of "voice cloud" by developers or hardware manufacturers widely used in a variety of applications and mobile phones and other devices, it will be the mobile internet revolution. In the existing mobile Internet applications, we have to see similar applications such as Airi, Smart 360, etc., which have through the voice of the way to open the application, query the weather and even "flirt" function, but the effect of recognition and answer is not ideal. We look forward to an official version of "Natural language understanding" in the end.

Personalized speech recognition

Personalized Today, voice can also play personalized. In the existing voice cloud platform, flying only provides the standard Mandarin recognition, for the accent of the dialect, there is no good recognition effect. Personalized speech recognition may provide a solution for dialect recognition. It is reported that the new generation of "voice cloud" platform can provide Shaanxi dialect, Sichuan dialect, northeast dialect and Cantonese speech recognition program. In fact, in the existing voice cloud platform already has the Northeast dialect and the Cantonese synthesis function--Saipan news broadcasts the multilingual broadcasting function. In addition to the dialect of personalized speech recognition, the next generation of "voice cloud" platform will also build a personal voice database, the database collects individual specific pronunciation information, and can quickly identify a language with a personal accent, that is, if it does not recognize Shanghainese, you can train and instruct it to learn to "listen" to Shanghai dialect.

Personalized speech synthesis

is similar to personalized speech recognition to synthesize different dialects. This is already implemented in the existing voice cloud platform. Further personalization is the synthesis of voice with personal characteristics. Since it is possible to identify individual features by establishing a personal voice database, believe that the synthesis of personal characteristics of the voice is not a problem, perhaps one day you do not want to speak, you can find a "double" for you, maybe one day when you call your friend, the other side has become Andy Lau or ge ... The meaning of personalized speech synthesis is rich!

Voice print recognition

in movies and TV dramas, we can often see the situation: The big guy came to the insurance door, the password, the Palm Press fingerprint recognition, green light sweep, voice requirements speak verification; the door's open! Such a scene is not a dream in the future. As the only non-contact personal characteristic, voice print has important significance in the application of public security. Perhaps one day, the police solve the case can not collect fingerprints; start the car without the key;Open the safe also do not have to set the password ..., the new generation of "voice cloud" platform of speech recognition characteristics, for various applications encryption provides a new generation of solutions.

Spoken evaluation

At present, the National Putonghua Proficiency test is basically used by the Hkust voice evaluation system. The application of oral evaluation function in life is very mature. In the next generation of "voice cloud" platform, oral evaluation has been opened, you can evaluate their Mandarin anytime, anywhere, when you are in Mandarin test, you can practice at home, master the rules, grasp the characteristics of the victorious. The

is the leader in the Chinese speech industry, the hkust, which has been ahead of the international it giants in smart voice, especially Chinese speech. Its 2010 release of the first "voice cloud" platform Products won the industry acclaim, the current flying voice Cloud development partners have reached more than 2,500, the number of end users reached 25 million, the number of online users more than 1.2 million per day, the total request for more than 7 million times. The release of a new generation of "voice cloud" platform is bound to be a major event in the development of phonetics.

(editor: Heritage)