维吾尔语语音语料库管理软件的研究与实现

摘要

Speech corpus is fundamental to both speech recognition and speech synthesis. Because of the characteristics of Uyghur language, there are a huge workload and complexity in building and managing of Uyghur speech corpus, this is the main reason for developing a Uyghur Speech Corpus MIS . In this paper we mainly discuss the design of Uyghur speech corpus DB which includes the design, REC and marking of speech. Then we discuss the overall design and key techniques in Uyghur Speech Corpus MIS.%语音语料库是语音识别和语音合成技术研究的基础.由于维吾尔语文字与语音特征、维吾尔语语音语料库的建立、管理、使用之工作不仅工作量巨大,且具有一定的复杂性,这就不得不开发维吾尔语语音数据库管理软件.本文首先研究维吾尔语语音语料库的设计过程,包括语音文本的设计、语音录制、语音库的标注、并在此基础上论述维吾尔语语音语料管理软件的总体功能设计以及部分关键技术的实现方法.