CASIA OpenIR  > 智能交互
基于静音时长和文本特征融合的韵律边界自动标注
傅睿博1,3; 李雅1; 温正棋1; 陶建华1,2,3
2017-10
Conference NameNCMMSC2017
Conference Date2017-10
Conference Place江苏连云港
PublisherNCMMSC2017
Abstract

韵律边界标注对于语料库建设和语音合成有着至关重要的作用,而自动韵律标注可以克服人工标注中的不一致、耗时的缺点。仿照人工标注流程,本文运用循环神经网络分别对文本和音频两个通道训练子模型,对子模型的输出采用模型融合,从而获得最优标注。我们以词为单位提取了静音时长,与传统以帧为单位的声学特征相比更加具有明确的物理意义,与韵律边界的联系更加紧密。实验结果表明,本文所采用的静音时长特征相比于传统声学特征对自动韵律标注的性能有所提高,决策融合方法相比于直接特征层面融合更好地结合了声学和文本的特征,进一步提高了标注的性能。

Language中文
Document Type会议论文
Identifierhttp://ir.ia.ac.cn/handle/173211/39599
Collection智能交互
Corresponding Author傅睿博
Affiliation1.中国科学院自动化研究所 模式识别国家重点实验室
2.中国科学院大学 人工智能技术学院
3.中国科学院自动化研究所 中国科学院脑科学与智能技术研究中心
First Author AffilicationChinese Acad Sci, Inst Automat, Natl Lab Pattern Recognit, Beijing 100190, Peoples R China;  Institute of Automation, Chinese Academy of Sciences
Corresponding Author AffilicationChinese Acad Sci, Inst Automat, Natl Lab Pattern Recognit, Beijing 100190, Peoples R China;  Institute of Automation, Chinese Academy of Sciences
Recommended Citation
GB/T 7714
傅睿博,李雅,温正棋,等. 基于静音时长和文本特征融合的韵律边界自动标注[C]:NCMMSC2017,2017.
Files in This Item: Download All
File Name/Size DocType Version Access License
NCMMSC2017(最佳学生论文).p(877KB)会议论文 开放获取CC BY-NC-SAView Download
Related Services
Recommend this item
Bookmark
Usage statistics
Export to Endnote
Google Scholar
Similar articles in Google Scholar
[傅睿博]'s Articles
[李雅]'s Articles
[温正棋]'s Articles
Baidu academic
Similar articles in Baidu academic
[傅睿博]'s Articles
[李雅]'s Articles
[温正棋]'s Articles
Bing Scholar
Similar articles in Bing Scholar
[傅睿博]'s Articles
[李雅]'s Articles
[温正棋]'s Articles
Terms of Use
No data!
Social Bookmark/Share
File name: NCMMSC2017(最佳学生论文).pdf
Format: Adobe PDF
All comments (0)
No comment.
 

Items in the repository are protected by copyright, with all rights reserved, unless otherwise indicated.