作者投稿和查稿 主编审稿 专家审稿 编委审稿 远程编辑

计算机工程 ›› 2011, Vol. 37 ›› Issue (20): 208-210. doi: 10.3969/j.issn.1000-3428.2011.20.072

• 人工智能及识别技术 • 上一篇    下一篇

基于子字单元的维吾尔语语音识别研究

薛化建 1,2,董兴华 1,2,周 喜 1,吐尔洪•吾司曼 1,李 晓 1   

  1. (1. 中国科学院新疆理化技术研究所,乌鲁木齐 830011;2. 中国科学院研究生院,北京 100190)
  • 收稿日期:2011-03-30 出版日期:2011-10-20 发布日期:2011-10-20
  • 作者简介:薛化建(1979-),男,博士研究生,主研方向:自然语言处理,语音识别;董兴华,博士研究生;周 喜,副研究员、博士研究生;吐尔洪?吾司曼,硕士;李 晓,研究员、博士生导师
  • 基金资助:
    中国科学院西部行动计划高新技术基金资助项目(KGC X2-YW-507)

Research on Uyghur Speech Recognition Based on Subword Unit

XUE Hua-jian 1,2, DONG Xing-hua 1,2, ZHOU Xi 1, Turghun Osman 1, LI Xiao 1   

  1. (1. Xinjiang Technical Institute of Physics and Chemistry, Chinese Academy of Sciences, Urumqi 830011, China; 2. Graduate University of Chinese Academy of Sciences, Beijing 100190, China)
  • Received:2011-03-30 Online:2011-10-20 Published:2011-10-20

摘要: 为提高维吾尔语语音识别的识别率,在分析维吾尔语特点的基础上,设计一种基于子字单元的维吾尔语语音识别总体结构,指出维吾尔语单词的发音模型,给出构建子字发音字典的方法,及其以子字单元为基础构建语言模型与声学模型的方法。在一个语音库上进行实验,采用一种非监督的词切分方法对维吾尔语单词进行词切分,生成子字。实验结果表明,基于子字单元的维吾尔语语音识别可以获得更好的识别结果。

关键词: 维吾尔语, 词切分, 子字单元, 隐马尔科夫模型, 连续语音识别

Abstract: To improve on accuracy of Uyghur speech recognition, based on analysis of Uyghur characteristics, the framework of Uyghur speech recognition based on subword is developed for the first time. Pronunciation model of Uyghur word is given. How to build subword pronouncing dictionary, subword language model and acoustic model is described. Experiments are completed on a speech corpus and an unsupervised Uyghur word segmentation method is utilized to produce subwords. Experimental results show that Uyghur speech recognition based on subword can gain better recognition results.

Key words: Uyghur, word segmentation, subword unit, Hidden Markov Model(HMM), continuous speech recognition

中图分类号: