支持重音合成的汉语语音合成系统

摘要
参考文献
相关文章 (15)

全文: PDF (2607 KB) HTML (1 KB)
输出: BibTeX | EndNote (RIS)

摘要针对基于单元挑选的汉语语音合成系统中重音预测及实现,本文采用了知识指导下的数据驱动建模策略。首先,采用经过感知结果优化的重音检测器,实现了语音数据库的自动标注;其次,利用重音标注数据库,训练得到支持重音预测的韵律预测模型;用重音韵律预测模型替代原语音合成系统中的相应模型,从而构成了支持重音合成的语音合成系统。实验结果分析表明,基于感知结果优化的重音检测器的标注结果是可靠的;支持重音的韵律声学预测模型是合理的;新的合成系统能够合成出带有轻重变化的语音。

	服务

	把本文推荐给朋友
	加入我的书架
	加入引用管理器
	E-mail Alert
	RSS
	作者相关文章
	朱维彬

关键词 ：计算机应用, 中文信息处理, 重音, 韵律模型, 语音合成

Abstract：To aim to predict and realize Chinese accent in a unit-selection based speech synthesis system, a data-driven method was used to build an accent-supported prosody module. First, with the help of Accent-Index detector which had been optimized with perceptual annotations, a speech corpus had been auto-annotated with Accent-Index. Then, a prosody predictive module supporting accent had been trained with the corpus. Replaced with the new prosody predictive module, the speech synthesis system could synthesize speech with various levels of accent. The results on the experiments had proved the accuracy of auto-detected accents, and the validity of the prosody predictor, and also the capability of accent realizing of the speech synthesis system.

Key words： computer application Chinese information processing accent prosodic model speech synthesis

收稿日期: 2006-12-25

中图分类号:

TP391

作者简介: : 朱维彬(1966—),男,博士,研究方向为语音声学模型,韵律建模以及言语数据库构建等。

链接本文:

http://jcip.cipsc.org.cn/CN/Y2007/V21/I3/122

[1] Ma Xijun, Zhang Wei, Zhu Weibin, et al. Probability Prosody Model for Unit Selection [A]. In: Proc. of ICASSP 2004 [C]. Montreal, Canada, 2004. 649-652.
[2] Chu Min, Peng Hu, et al. Selecting Non-Uniform Units from a Very Large Corpus for Concatenative Speech Synthesizer [A]. In: Proc. of ICASSP 2001 [C]. Sale Lake City, USA, 2001.
[3] 初敏. 自然言语的韵律组织中的不确定性及其在语音合成中的应用 [J]. 中文信息学报,2004,18(4): 66-71.
[4] 陶建华,赵晟,蔡莲红. 基于统计韵律模型的汉语语音合成系统的研究 [J]. 中文信息学报,2002,16(4): 1-6.
[5] 吴志勇,蔡莲红. 语音合成中的韵律关联模型 [J]. 中文信息学报,18(2): 44-50.
[6] 吴晓如,王仁华,刘庆峰. 基于韵律特征和语法信息的韵律边界检测模型 [J]. 中文信息学报,2003,17(5): 48-54.
[7] Chen Gaopeng, Hu Yu, Wang Renhua, et al. Quantitative Analysis and Synthesis of Focus in Mandarin [A]. In: Proc. of TAL 2004 [C]. Beijing, China, 2004. 25-28.
[8] Greg Kochanski, Chilin Shih, et al. Hierarchical Structure and Word Strength Prediction of Mandarin Prosody [J]. International Journal of Speech Technology, 2003,6(1): 33-43.
[9] 王韫佳,初敏,贺琳. 普通话语句重音在双音节韵律词中的分布 [J]. 语言科学,2004,3(5): 38-48.
[10] Zhu Weibin, Shi Qin, et al. Corpus Building for Data-Driven TTS Systems [A]. In: Proc. of IEEE TTS Workshop 2002 [C]. Santa Monica, USA, 2002. 199-202.
[11] Zhu Weibin, Zhang Wei, Shi Qin, et al. Automatic Detection of Chinese Accent-Index Based on Approximation-Ratio [A]. In: Proc. of ISCSLP 2004 [C]. Hong Kong, China, 2004. 85-88.
[12] Zhu Weibin. Perceptual Optimization of the Chinese Accent-Index Detector [A]. In: Proc. of Speech Prosody 2006 [C]. Dresden, German, 2006.
[13] Xu Yi. Effects of tone and focus on the formation and alignment of f0 contours [J]. Journal of Phonetics, 1999, 27: 55-105.
[14] Shi Qin, Ma Xijun, Zhu Weibin, et al. Statistic Prosody Structure Prediction [A]. In: Proc. of IEEE TTS Workshop 2002 [C]. Santa Monica, USA, 2002.