Investigating Deep Neural Network Adaptation for Generating Exclamatory and Interrogative Speech in Mandarin | |
Zheng, Yibin1,3; Li, Ya1; Wen, Zhengqi1; Liu, Bin1; Tao, Jianhua1,2,3; Jianhua Tao | |
刊名 | JOURNAL OF SIGNAL PROCESSING SYSTEMS FOR SIGNAL IMAGE AND VIDEO TECHNOLOGY |
2018-07-01 | |
卷号 | 90期号:7页码:1039-1052 |
关键词 | Speech Synthesis Excitation Parameters Deep Neural Network Adaptation Exclamatory Speech Interrogative Speech |
DOI | 10.1007/s11265-017-1290-2 |
文献子类 | Article |
英文摘要 | Currently, most speech synthesis systems only generate speech in a reading style, which greatly affects the expressiveness of the synthetized speech. To improve the expressiveness of the synthetized speech, this paper focuses on the generation of exclamatory and interrogative speech for Mandarin spoken language. A multi-style (exclamatory and interrogative) deep neural network-based acoustic model with a style-specific layer (which can have multiple layers) and several shared hidden layers is proposed. The style-specific layer is used to model the distinct style specific patterns. The shared layers allow maximum knowledge sharing between the declarative and multi-style speech. We investigate five major aspects of the multi-style adaptation: neural network type and topology, the number of layers in style-specific layer, initial model, adaptation parameters and adaptation corpus size. Both objective and subjective evaluations are carried out to evaluate the proposed method. Experiment results show the proposed multi-style BLSTM with top one layer adapted is superior to our prior work (which is trained by the combination of constrained Maximum likelihood linear regression and structural maximum a posterior), and achieves the best performance. We also find that adapting on both spectral and excitation parameters are more effective than only adapting on the excitation parameters. |
WOS关键词 | Short-term-memory ; Speaker Adaptation ; Synthesis System ; Emotional Expressions ; Model ; Algorithms ; Features ; Styles ; Pitch ; Hsmm |
WOS研究方向 | Computer Science ; Engineering |
语种 | 英语 |
WOS记录号 | WOS:000433555600008 |
资助机构 | National High-Tech Research and Development Program of China (863 Program)(2015AA016305) ; National Natural Science Foundation of China (NSFC)(61305003 ; Strategic Priority Research Program of the CAS(XDB02080006) ; Major Program for the National Social Science Fund of China(13ZD189) ; 61425017 ; 61403386) |
内容类型 | 期刊论文 |
源URL | [http://ir.ia.ac.cn/handle/173211/19885] |
专题 | 自动化研究所_模式识别国家重点实验室_人机语音交互团队 |
通讯作者 | Jianhua Tao |
作者单位 | 1.Chinese Acad Sci Recognit, Inst Automat, Natl Lab Pattern Recognit, Beijing, Peoples R China 2.Chinese Acad Sci, Inst Automat, CAS Ctr Excellence Brain Sci & Intelligence Techn, Beijing, Peoples R China 3.Univ Chinese Acad Sci, Sch Artificial Intelligence, Beijing, Peoples R China |
推荐引用方式 GB/T 7714 | Zheng, Yibin,Li, Ya,Wen, Zhengqi,et al. Investigating Deep Neural Network Adaptation for Generating Exclamatory and Interrogative Speech in Mandarin[J]. JOURNAL OF SIGNAL PROCESSING SYSTEMS FOR SIGNAL IMAGE AND VIDEO TECHNOLOGY,2018,90(7):1039-1052. |
APA | Zheng, Yibin,Li, Ya,Wen, Zhengqi,Liu, Bin,Tao, Jianhua,&Jianhua Tao.(2018).Investigating Deep Neural Network Adaptation for Generating Exclamatory and Interrogative Speech in Mandarin.JOURNAL OF SIGNAL PROCESSING SYSTEMS FOR SIGNAL IMAGE AND VIDEO TECHNOLOGY,90(7),1039-1052. |
MLA | Zheng, Yibin,et al."Investigating Deep Neural Network Adaptation for Generating Exclamatory and Interrogative Speech in Mandarin".JOURNAL OF SIGNAL PROCESSING SYSTEMS FOR SIGNAL IMAGE AND VIDEO TECHNOLOGY 90.7(2018):1039-1052. |
个性服务 |
查看访问统计 |
相关权益政策 |
暂无数据 |
收藏/分享 |
除非特别说明,本系统中所有内容都受版权保护,并保留所有权利。
修改评论