根据语音合成与识别等语音应用研究的需求,从文本分析模块入手,利用"维吾尔语语音声学参数库",选择了包含辅音/r/的单音节以及多音节词,对其语图中弱短元音现象进行以实验语音学为出发点的研究。对声学参数进行统计,归纳其共振峰、音强和时长分布模式,得出了一系列结论,其目的是为了提高语音合成的自然度,即更好的为自然语言处理服务。本项研究结果对维吾尔语语言乃至整个阿尔泰语系语言的韵律研究具有较高的参考价值。
Based on the needs of speech processing applications,such as the improvement of speech synthesis naturalness,and based on the "Uyghur Acoustical feature Database" containing acoustical measurements of Consonant /r/ in words by 1 male and 1 female speakers separately,the analysis results on the systematic empirical acoustical features of the consonant /r/ in Uyghur language are presented,and lay a foundation for studying the prosody in words and sentences of Uyghur language.These research results are of high value for further studying the technology of Uyghur natural language processing,and also serve as a valuable reference for the study of entire Altay language family.