一个面向言语工程的兰州方言语料库  

A Lanzhou dialect corpus for speech engineering

作　　者：杨鸿武[1] 梁青青[1] 郭威彤[1] 李锦珑[1] 陈龙[1] 

机构地区：[1]西北师范大学物理与电子工程学院,甘肃兰州730070

出　　处：《西北师范大学学报（自然科学版）》2009年第6期54-59,共6页Journal of Northwest Normal University(Natural Science)

基　　金：国家自然科学基金面上项目(60875015);教育部科学研究重点项目(208146)

摘　　要：建立了一个兰州方言语料库.利用《方言调查字表》设计了兰州方言的文本语料,包括1 280个单字,2 000个双字词,18个负载句和1 000句语句,录制了4位发音人的兰州方言和普通话平行的语音语料,并标注了音节边界、峰值点、声韵母和声调信息;对于语句,根据词法信息和语法信息,利用TBL算法标注了韵律边界.语料库可用于兰州方言的实验语音学研究以及兰州方言的韵律建模、语音合成、语音转换的研究.This paper built a Lanzhou dialect corpus with text and speech.Text corpus,which includes 1 280 single syllables,2 000 two-syllable words,18 carrier sentences and 1 000 sentences,is designed based on＂Dialect word list＂.Speech corpus is recorded by four local speakers with contrastive（Lanzhou dialect vs.Mandarin） recordings.Syllable boundaries,peak points,initial,final and tone information are labeled for each syllable of speech corpus.The prosodic boundaries of the text sentences are labeled by TBL algorithm with grammatical information and lexical information. The corpus can be applied to study experimental phonetics of Lanzhou dialect and prosody modeling, text-to-speech synthesis as well as voice conversion for Lanzhou dialect.

关 键 词：兰州方言 语料库 语音切分 峰值点标注 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]