学习型中文语音识别系统研究及实现

由于中文存在多音多义字,所以相同的发音可能代表不同的含义。而每个人的发音又会有一定的差异。又因为地方方言等原因,导致中文语音控制的正确性或者成功正确地识别变得很困难。采用逆向思维,通过系统记录使用人的发音,并把使用人自己设定的发音绑定到相应的功能或功能命令,以实现学习型的中文语音识别。

The Learnling Style of Chinese Speech Recognition System's Research and Implementation

As the Chinese presence multiple-reading characters,the same pronunciation may represent different meanings.And everyone has some differences in pronunciation.The local dialect and other reasons led to the correctness of the Chinese voice control or correctly identify becomes very difficult.In this paper,the use of reverse thinking,we recorded the pronunciation of users by the system,and set their own pronunciation bound to the corresponding function or functions of command,in order to achieve the learning of Chinese speech recognition.