基于微格矢量化编码与缺陷跟踪的汉语语音数据采集研究  

A Study on Data Collecting of Chinese Pronunciation of Lao Students Based on VQF and DT Mechanism

作　　者：陈展[1] 

机构地区：[1]昆明理工大学国际学院,昆明650093

出　　处：《价值工程》2016年第2期149-151,共3页Value Engineering

基　　金：云南省教育厅本科教学质量与教学改革工程项目-"东南亚语种人才培养示范点";项目代码:10968258

摘　　要：本文以老挝留学生为例,阐述了采集其汉语语音数据的基本原则和基本过程,同时提出了一种使用微格VQF音频技术(Transform-domain Weighted I Nterleave Vector Quantization)和缺陷跟踪机制(Defect Tracking Mechanism)来采集数据的方法,避免了传统语音数据采集过程中的弊端,使采集到的语音数据准确、规范,从而提高老挝学生汉语语音习得的效率,为建立东南亚留学生汉语语音库奠定了坚实的数据基础。Depending on the Chinese pronunciation of Lao students,the article demonstrates the fundamental principals and process of data collecting on pronunciation performance, furthermore introduces a method which collects evaluation data through VQF technique in Micro Teaching and the defect-tracking mechanism. It has challenged the present process of data collecting of Chinese pronunciation. As a result, the collected data with accuracy and regularity will be gained and the efficiency of Lao′ s pronunciation learning will be improved,and the data base of Chinese pronunciation for foreign students from southeast of Asia will be built.

关 键 词：矢量化编码 缺陷跟踪 语音 数据采集 

分 类 号：O157.4[理学—数学]