[1]张永皋,马青玉,孙 青.基于MFCC和CHMM技术的语音情感分析及其在教育中的应用研究[J].南京师范大学学报(工程技术版),2009,09(02):089-92.
　Zhang Yonggao,Ma Qingyu,Sun Qing.Investigation on Speech Emotion Analyses and Its Application in Education Based on MFCC and CHMM Techniques[J].Journal of Nanjing Normal University(Engineering and Technology),2009,09(02):089-92.
点击复制

基于MFCC和CHMM技术的语音情感分析及其在教育中的应用研究
分享到：

南京师范大学学报（工程技术版）[ISSN:1006-6977/CN:61-1281/TN]

卷:
09卷
期数:
2009年02期
页码:
089-92
栏目:
出版日期:
2009-06-30

文章信息/Info

Title:
Investigation on Speech Emotion Analyses and Its Application in Education Based on MFCC and CHMM Techniques
作者:
张永皋1;  马青玉2;  孙 青1
1. 南京师范大学教育科学学院, 江苏南京210097;
2. 南京师范大学物理科学与技术学院, 江苏南京210097
Author(s):
Zhang Yonggao1; Ma Qingyu2; Sun Qing1
1.School of Educational Science,Nanjing Normal University,Nanjing 210097,China;2.School of Physics and Technology,Nanjing Normal University,Nanjing 210097,China
关键词:
MFCC;  CHMM;  情感识别
Keywords:
MFCC;  CHMM;  speech emo tion recognition
分类号:
TN912.34
摘要:
语音情感识别作为一个新的研究热点,因其能解决教育中情感缺失的问题,而越来越受到研究者的重视.选取符合人类听觉系统感知的M el频率倒谱系数(MFCC)与各态历经型的连续隐马尔可夫模型(CHMM)进行语音情感特征的分析,并对大量的语音信号进行情感识别实验,识别正确率达到86.7%,为教育中的情感补偿提供了切实可行的依据.
Abstract:
A s a new research ho tspo t, speech emo tion recogn ition attractsm ore and atten tions from researchers fo r its solu tion o f the em otion lo ss in d istance education. Based on the perception o f the hum an auditory system, theM e l Frequency C epstral Coe fficients (M FCC) and the ContinuousH iddenM arkovM ode l ( CHMM ) techn iques are used in this paper fo r speech charac teristic ana lyses and em otion identifica tion. The recogn ition accuracy o f 871 6% is ach ieved in the expe rim ent for large num bers of speech s igna ls and the potentia l application for em otion com pensation is prov ided in education.

参考文献/References:

[ 1] 张迎辉, 林学誾. 情感可以计算) ) ) 情感计算综述[ J]. 计算机科学, 2008, 35( 5): 5-8.
Zhang Y inghu,i Lin Xuey in. A ffect is com putable) ) ) a survey on a ffective com puting[ J]. Com pute r Science, 2008, 35( 5):
5-8. ( in Ch inese)
[ 2] 隋云翔, 刘平. 非言语沟通与课堂教学[ J]. 现代中小学教育, 1991( 4) : 50-52.
Sui Yunx iang, Liu P ing. Non-verba l comm un ication in classroom teach ing [ J] . M odern Pr im ary and Secondary Educa tion,
1991( 4): 50-52. ( in Chinese)
[ 3] 蒋丹宁, 蔡莲红. 基于语音声学特征的情感信息识别[ J]. 清华大学学报: 自然科学版, 2006, 46( 1): 86-89.
Jiang Dann ing, Ca iL ianhong. Speech em o tion recogn ition using acoustic features[ J]. Journal of TsinghuaUn iversity: Science
and Technology Edition, 2006, 46( 1): 86-89. ( in Ch inese)
[ 4] 詹永照, 曹鹏. 语音情感特征提取和识别的研究与实现[ J]. 江苏大学学报: 自然科学版, 2005, 26( 1): 72-75.
Zhan Yongzhao, Cao Peng. Research and im plem entation of emo tional fea ture ex traction and recognition in speech signa l[ J].
Jou rnal of Jiang su Un iversity: Na tura l Sc ience Ed ition, 2005, 26( 1) : 72-75. ( in Chinese)
[ 5] 赵力, 将春辉, 邹采荣, 等. 语音信号中的情感特征分析和识别的研究[ J]. 电子学报, 2004, 32( 4): 606-609.
Zhao L,i Jiang Chunhu,i Zou Ca irong, et a .l A study on em otiona l feature ana lysis and recognition in speech[ J] . Acta E lectron
ica S inica, 2004, 32( 4): 606-609. ( in Chinese)
[ 6] Schu ller B, R igo ll G, LangM. H iddenM arkovm ode-l based speech em o tion recogn ition [ C] / / IEEE Internationa lConference
on Acoustics, Speech, & Signa l Processing. H ongkong, China, 2003: 401-404.
[ 7] 赵力. 语音信号处理[M ]. 北京: 机械工业出版社, 2003.
Zhao L.i Vo ice S igna l Processing[M ]. Be ijing: Ch inaM ach ine Press, 2003. ( in Chinese)
[ 8] AyadiM, K am elM, K array F. Speech em otion recogn ition using Gaussian m ix ture vecto r autoregressivem ode ls [ C] / / IEEE
Interna tiona l Conference on Acoustics, Speech, & S ignal Processing. H ono lu lu, H aw a i,i USA, 2007: IV - 957- IV - 960.
[ 9] de CheveignA, K aw aharaH. YIN, a fundamenta l frequency estim ator fo r speech and m us ic [ J]. Journa lA coustica l So ciety
o f Am er ica, 2002, 111( 4): 1 917-1 930.
[ 10] Nogue iras A, M oreno A, Bona fonte A, et a .l Speech em o tion recognition us ing h iddenM arkov mode ls[ C ] / / Pro ceedings o f
Euro speech. Aalbo rg, Denm ark, 2001: 2 679-2 682.

备注/Memo

备注/Memo:
基金项目: 教育部留学回国人员科研启动基金( 2008102SB J0154)、南京师范大学留学回国人员基金( 2008102XLH 0070 )和南京师范大学
教务处精品实验资助项目.
通讯联系人: 马青玉, 博士, 副教授, 研究方向: 声信号处理与生物医学电子技术应用. E-m ail:m aqingyu@ n jnu. edu. cn
更新日期/Last Update: 2013-04-23