赵云雪，张珑，郑世杰.短时谱特征的汉语重音检测方法研究[J].计算机科学与探索，2014，8（9）：1120-1128.
ISSN 1673-9 418 CODEN JKYTA8
Journal of Frontiers of Computer Science and Technology
1673-9 418/2014/08( 09)-1120-09
doi: 1 0.3778/j.issn.1673-9418.1407004
E-mail: fcst@vip.1 63.co m
http://www.ceaj.org
Tel: +86-10-89056056
短时谱特征的汉语重音检测方法研究
*
赵云雪
1
，张 珑
1,2+
，郑世杰
1
1. 哈尔滨师范大学 计算机科学与信息工程学院，哈尔滨 150025
2. 哈尔滨工业大学 计算机科学与技术学院，哈尔滨 1500 01
Chinese Accent Detection Method Research Based on Short-Time Spectrum Features
ZH AO Yunxue
1
, ZHANG Long
1,2+
, ZHENG Shijie
1
1. College of Computer Science and Information Engi ne ering, Harbin Normal University, Harbin 150025, C hina
2. School of Computer Science an d Technology, Harbin Institute of Technology, Harbin 150001, China
+ Cor responding author: E-mail: zlwalkman@sina.com
ZHAO Yunxue, ZHANG Long, ZHENG Shijie. Chinese accent detection method research based on short-time
spectrum features. Journal of Frontiers of Computer Science and Technology, 2014, 8(9)：1120-1128.
Abstr ac t: Accent is a critically important component of spoken communication, and plays a very i mportant role in
spoken communicatio n. In o rder to veri fy the effect of s hort-time spectrum feature set based on auditory model in
Chinese accent detection method, this paper uses MFCC (Mel frequency cepstrum coefficient) algorithm and RASTA-
PLP (relative spectra perceptual linear prediction) algorithm to extract each v oice segment of short-time spectrum
information, and builds short-time spectrum feature sets based on MFCC a lgorithm and RASTA-PLP algorithm.
T he n, it chooses NaiveBayes classif ier to model the two feat ure sets, and chooses the classes with maximum a poste-
ri ori pro ba bility a s the object’s class. This classification method makes full use of th e related phonet ic features of
speech segment. Sh ort-time spectrum feature set based on MFCC and short-time spectrum feature set based on RASTA-
PLP respectively achieve 82.1% and 80.8% accent detection accuracy on ASCCD (annotated speech corpus of Chi-
nese discourse). The experimental results indicate that short-time spectrum features based on MFCC and short-time
spectrum features based on RASTA - PLP can be used for Chinese accent detection resear ch .
* The Natural Science Foundation of Heil ongjiang Province of China under Grant No. F2 01321 (黑龙江省自然科学基金); the Philoso-
phy an d Social Sciences Foreign Language Joint Research Foundation of Heilongjiang Province of China under G rant No. 12H007
(黑龙江省哲学社会科学外语联合研究项目).
Received 2014-06, Accepted 2014-08.
CNKI网络优先出版：2014-08-19, http://www.cnki.net/kcms/doi/10.3778/j.issn.1 673-9418.1407004.html
下载后可阅读完整内容，剩余8页未读，立即下载
weixin_38572115
- 粉丝: 6
- 资源: 946
我的内容管理
展开
最新资源
- 解决Eclipse配置与导入Java工程常见问题
- 真空发生器：工作原理与抽吸性能分析
- 爱立信RBS6201开站流程详解
- 电脑开机声音解析：故障诊断指南
- JAVA实现贪吃蛇游戏
- 模糊神经网络实现与自学习能力探索
- PID型模糊神经网络控制器设计与学习算法
- 模糊神经网络在自适应PID控制器中的应用
- C++实现的学生成绩管理系统设计
- 802.1D STP 实现与优化：二层交换机中的生成树协议
- 解决Windows无法完成SD卡格式化的九种方法
- 软件测试方法：Beta与Alpha测试详解
- 软件测试周期详解：从需求分析到维护测试
- CMMI模型详解：软件企业能力提升的关键
- 移动Web开发框架选择：jQueryMobile、jQTouch、SenchaTouch对比
- Java程序设计试题与复习指南
资源上传下载、课程学习等过程中有任何疑问或建议，欢迎提出宝贵意见哦~我们会及时处理！
点击此处反馈
