正在加载图片...
第11卷第2期 智能系统学报 Vol.11 No.2 2016年4月 CAAI Transactions on Intelligent Systems Apr.2016 D0I:10.11992/is.201511008 网络出版地址:http://www.enki..net/kcms/detail/23.1538.TP.20160315.1248.018.html 一种语音特征提取中Ml倒谱系数的后处理算法 张毅,谢延义2,罗元3,席兵3 (1.重庆邮电大学先进制造工程学院,重庆400065:2.重庆邮电大学自动化学院,重庆400065:3.重庆邮电大学光 电工程学院,重庆400065) 摘要:为提高语音识别系统的鲁棒性,本文以Ml频率倒谱系数(MFCC)为基础,结合均值消减法、方差归一化、时 间序列滤波法和加权自回归移动平均滤波法,提出了一种后处理算法,本文将该算法命名为MVDA后处理法,所得 语音特征参数简称MVDA。本文首先从理论上推导了MVDA后处理法可以去除加性噪声和卷积噪声的干扰,接着 针对MVDA与MFCC做了对比试验,并分析了含噪语音与语音信号的欧氏距离变化,证明MVDA后处理法的每一步 均有效降低了噪声的干扰,且得出了MVDA在不同噪声环境中均更优的结论。这种简洁的语音特征不仅可以达到 许多复杂语音特征处理方法的效果,而且有效减少了自动语音识别系统的计算量。 关键词:后处理:语音特征:语音识别:噪声:鲁棒性 中图分类号:TP391.4文献标志码:A文章编号:1673-4785(2016)02-0208-07 中文引用格式:张毅,谢延义,罗元,等.一种语音特征提取中Ml倒谱系数的后处理算法[J].智能系统学报,2016,11(2): 208-215. 英文引用格式:ZHANG Yi,XIE Yanyi,LUO Yuan,etal.Postprocessing method of MFCC in speech feature extraction[J].CAAI transactions on intelligent systems,2016,11(2):208-215. Postprocessing method of MFCC in speech feature extraction ZHANG Yi',XIE Yanyi2,LUO Yuan',XI Bing' (1.Institute of Advanced Manufacturing Engineering,Chongqing University of Posts and Telecommunications,Chongqing 400065, China;2.College of Automation,Chongqing University of Posts and Telecommunications,Chongqing 400065,China;3.College of Opto Electronic Engineering,Chongqing University of Posts and Telecommunications,Chongqing 400065,China) Abstract:To improve the robustness of automatic speech recognition systems,a new speech feature postprocessing method based on the Mel-frequency Cepstral Coefficient MFCC)is proposed,which is named the MVDA postpro- cessing method.The postprocessed feature parameters are named MVDAs.This technique combines mean subtrac- tion,variance normalization,time sequence fltering,and autoregressive moving average flters.Experiments were conducted to compare MVDA and MFCC.Changes in the Euclidean distance of the speech with noise and the speech signal were analyzed,proving that every step of MVDA postprocessing could effectively reduce the noise in- terference.Thus,all MVDAs in different noise environments were superior.This simple feature does not only a- chieve the effect of many complex speech feature processing methods but also effectively reduces the computational complexity of automatic speech recognition systems. Keywords:postprocessing;phonetic feature;speech recognition;noise;robustness 为了提高语音识别系统的鲁棒性,谱减法、卡尔 收稿日期:2015-11-06.网络出版日期:2016-03-15. 曼滤波1]和麦克风阵列[]等语音增强技术得到应 基金项目:重庆市科委前沿技术专项重点项目(cstc2015 jeyjBX0066). 通信作者:谢延义.E-mail:811719530@qq.com. 用和推广。语音特征的失真造成声学空间的变形,第 11 卷第 2 期 智 能 系 统 学 报 Vol.11 №.2 2016 年 4 月 CAAI Transactions on Intelligent Systems Apr. 2016 DOI:10.11992 / tis.201511008 网络出版地址:http: / / www.cnki.net / kcms/ detail / 23.1538.TP.20160315.1248.018.html 一种语音特征提取中 Mel 倒谱系数的后处理算法 张毅1 ,谢延义2 ,罗元3 ,席兵3 (1.重庆邮电大学 先进制造工程学院,重庆 400065; 2. 重庆邮电大学 自动化学院,重庆 400065; 3. 重庆邮电大学 光 电工程学院,重庆 400065) 摘 要:为提高语音识别系统的鲁棒性,本文以 Mel 频率倒谱系数(MFCC)为基础,结合均值消减法、方差归一化、时 间序列滤波法和加权自回归移动平均滤波法,提出了一种后处理算法,本文将该算法命名为 MVDA 后处理法,所得 语音特征参数简称 MVDA。 本文首先从理论上推导了 MVDA 后处理法可以去除加性噪声和卷积噪声的干扰,接着 针对 MVDA 与 MFCC 做了对比试验,并分析了含噪语音与语音信号的欧氏距离变化,证明 MVDA 后处理法的每一步 均有效降低了噪声的干扰,且得出了 MVDA 在不同噪声环境中均更优的结论。 这种简洁的语音特征不仅可以达到 许多复杂语音特征处理方法的效果,而且有效减少了自动语音识别系统的计算量。 关键词:后处理;语音特征;语音识别;噪声;鲁棒性 中图分类号:TP391.4 文献标志码:A 文章编号:1673⁃4785(2016)02⁃0208⁃07 中文引用格式:张毅,谢延义,罗元,等. 一种语音特征提取中 Mel 倒谱系数的后处理算法[ J] . 智能系统学报, 2016, 11( 2) : 208⁃215. 英文引用格式:ZHANG Yi,XIE Yanyi,LUO Yuan, et al. Postprocessing method of MFCC in speech feature extraction[ J]. CAAI transactions on intelligent systems, 2016, 11(2): 208⁃215. Postprocessing method of MFCC in speech feature extraction ZHANG Yi 1 , XIE Yanyi 2 , LUO Yuan 3 , XI Bing 3 (1. Institute of Advanced Manufacturing Engineering, Chongqing University of Posts and Telecommunications, Chongqing 400065, China; 2. College of Automation, Chongqing University of Posts and Telecommunications, Chongqing 400065, China; 3. College of Opto Electronic Engineering, Chongqing University of Posts and Telecommunications, Chongqing 400065, China) Abstract:To improve the robustness of automatic speech recognition systems, a new speech feature postprocessing method based on the Mel⁃frequency Cepstral Coefficient (MFCC) is proposed, which is named the MVDA postpro⁃ cessing method. The postprocessed feature parameters are named MVDAs. This technique combines mean subtrac⁃ tion, variance normalization, time sequence fltering, and autoregressive moving average flters. Experiments were conducted to compare MVDA and MFCC. Changes in the Euclidean distance of the speech with noise and the speech signal were analyzed, proving that every step of MVDA postprocessing could effectively reduce the noise in⁃ terference. Thus, all MVDAs in different noise environments were superior. This simple feature does not only a⁃ chieve the effect of many complex speech feature processing methods but also effectively reduces the computational complexity of automatic speech recognition systems. Keywords: postprocessing; phonetic feature; speech recognition; noise; robustness 收稿日期:2015⁃11⁃06. 网络出版日期:2016⁃03⁃15. 基金项目:重庆市科委前沿技术专项重点项目(cstc2015jcyjBX0066). 通信作者:谢延义. E⁃mail:811719530@ qq.com. 为了提高语音识别系统的鲁棒性,谱减法、卡尔 曼滤波[1⁃2]和麦克风阵列[3] 等语音增强技术得到应 用和推广。 语音特征的失真造成声学空间的变形
向下翻页>>
©2008-现在 cucdc.com 高等教育资讯网 版权所有