基于SQLite技术的汉语语音识别数据库的建立

建立一个适合于特定说话人识别系统的汉语语音识别数据库，对推动说话人识别技术的研究和应用具有重要意义。本文基于支持向量机的说话人识别系统研究和开发过程中，构建了一个基于SQLite技术的汉语语音识别数据库，通过LabVIEW平台来实现对数据库控制操作。采用无序列样本和语音数据库样本分别进行比对实验。测试结果表明：一方面，无论是采用该语音识别数据库样本还是无序样本对说话人识别系统的识别率没有改变，这充分说明本系统建立的汉语语音识别数据库具有高稳定性和可靠性；另一方面，采用语音识别数据库样本其系统识别时间却明显缩短，这是改善基于支持向量机的说话人识别系统性能的有效途径。

Based on SQLite Technology Establishment Chinese Speech Recognition Database

For the establishment of a particular model of speaker recognition system Chinese speech recognition database, speech recognition technology on the promotion of research and application of great significance. SVM Based Speaker Recognition System research and development process, built a SQLite-based database Mandarin speech recognition technology, through the LabVIEW platform to implement control operations on the database. Sequence of sample-free and speech database to compare the experimental samples, respectively. The results show that: on the one hand, both the sample database using the speech recognition or speaker recognition disordered sample recognition rate did not change the system, which fully shows that the establishment of the Chinese speech recognition system with high stability and reliability of the database; other On the one hand, the sample database using speech recognition system to recognize their time is significantly reduced, which is to improve the SVM-based speaker recognition system performance effective way.