Please wait a minute...
 首页  期刊介绍 期刊订阅 联系我们 横山亮次奖 百年刊庆
 
最新录用  |  预出版  |  当期目录  |  过刊浏览  |  阅读排行  |  下载排行  |  引用排行  |  横山亮次奖  |  百年刊庆
清华大学学报(自然科学版)  2018, Vol. 58 Issue (6): 603-608    DOI: 10.16511/j.cnki.qhdxxb.2018.26.027
  电子工程 本期目录 | 过刊浏览 | 高级检索 |
基于波束形成与最大似然估计的近距离双麦克风语音增强算法
宫琴1,2, 郑硕1
1. 清华大学 生物医学工程系, 北京 100084;
2. 清华大学 深圳研究生院, 生物医学研究中心, 深圳 518055
Beamforming and maximum likelihood estimation for speech enhancement using dual closely-spaced microphones
GONG Qin1,2, ZHENG Shuo1
1. Department of Biomedical Engineering, Tsinghua University, Beijing 100084, China;
2. Research Center of Biomedical Engineering, Graduate School at Shenzhen, Tsinghua University, Shenzhen 518055, China
全文: PDF(2526 KB)  
输出: BibTeX | EndNote (RIS)      
摘要 为了解决波束形成法直接应用于近距离双麦克风系统时存在的问题,如目标信号低频段能量损失、多方向的竞争性语音噪声难以被有效抑制等,该文提出一种基于波束形成与最大似然估计的两步去噪方法。该方法首先使用加权叠加滤波器对混合声源进行时频分解,然后通过2个零点分别在0°和180°的波束形成图的幅频响应比值,设置各时频单元所对应的初步掩蔽值,在避免低频滚降现象出现的情况下,抑制本底噪声;最后根据统计模型和简化的最大似然估计法,抑制多方向的竞争性语音噪声,进一步增强目标信号。测试结果表明:在低信噪比、多种类型噪声源共同存在的情境下,该方法可以在无需低通滤波或宽带波束补偿的情况下,恢复原始信号的能量分布特点,明显提升信噪比。
服务
把本文推荐给朋友
加入引用管理器
E-mail Alert
RSS
作者相关文章
宫琴
郑硕
关键词 近距离双麦克风波束形成最大似然估计低频滚降多方向竞争性语音噪声    
Abstract:Traditional beamforming systems using dual closely-spaced microphones have various problems such as low-frequency-roll-off and limitations in suppressing competitive speech noises from multiple directions. This paper presents a two-step beamforming and maximum likelihood estimation algorithm. The algorithm first uses a WOLA filter for the time-frequency analysis for the speech mixture and then sets mask values to suppress background noise without low-frequency-roll-off based on the ratio of the two beamforming patterns, which have zeros at 0ånd 180°. A statistical model and the maximum likelihood estimation are then used to further enhance the speech. Tests indicate that the algorithm effectively recovers the energy distribution of the target signal and improves the signal-to-noise ratio without a low-pass filter or broadband compensation when the signal-to-ratio is low or multiple kinds of noises exist.
Key wordsclosely-spaced dual microphones    beamforming    maximum likelihood estimation    low-frequency roll-off    competitive speech noises from multiple directions
收稿日期: 2017-12-07      出版日期: 2018-06-21
基金资助:国家自然科学基金资助项目(61271133);高等学校博士学科点专项科研基金资助项目(20120002110054);清华信息科学与技术国家实验室(筹)学科交叉基金资助项目(042003172);深圳市自然科学基金资助项目(JCYJ20140827160046745)
引用本文:   
宫琴, 郑硕. 基于波束形成与最大似然估计的近距离双麦克风语音增强算法[J]. 清华大学学报(自然科学版), 2018, 58(6): 603-608.
GONG Qin, ZHENG Shuo. Beamforming and maximum likelihood estimation for speech enhancement using dual closely-spaced microphones. Journal of Tsinghua University(Science and Technology), 2018, 58(6): 603-608.
链接本文:  
http://jst.tsinghuajournals.com/CN/10.16511/j.cnki.qhdxxb.2018.26.027  或          http://jst.tsinghuajournals.com/CN/Y2018/V58/I6/603
  图1 声源传播示意图
  图2 两个波束形成系统输出幅度的比值 与声源来波方向的关系
  图3 幅频响应图
  图4 算法流程图
  图5 场景1的处理效果
  表1 3种信号在所选择的刺激通道的功率谱密度
  图6 场景2的处理效果
  表2 信号在所选刺激通道的功率谱密度
  表3 3种方法主要评价指标的对比
[1] NOH J, JO H, PARK Y, et al. Acoustic-focusing headphone based on delay-and-sum beamforming[C]//Proceedings of 2010 International Conference on Control, Automation and Systems. Gyeonggi-do, Korea:IEEE Press, 2010:2061-2064.
[2] KOKKINAKIS K, RUNGE C, TAHMINA Q, et al. Evaluation of a spectral subtraction strategy to suppress reverberant energy in cochlear implant devices[J]. Journal of the Acoustical Society of America, 2015, 138(1):115-124.
[3] SALEEM N. Single channel noise reduction system in low SNR[J]. International Journal of Speech Technology, 2017, 20(1):89-98.
[4] MAHIEUX Y, LE TOURNEUR G, SALIOU A. A microphone array for multimedia workstations[J]. Journal of the Audio Engineering Society, 1996, 44(5):365-372.
[5] BRANDSTEIN M S, WARD E D B. Microphone arrays:Signal processing techniques and applications[M]. Berlin:Springer, 2001.
[6] GRIFFITHS L J, JIM C W. An alternative approach to linearly constrained adaptive beamforming[J]. IEEE Transactions on Antenaas Propagation, 1982, 30(1):27-34.
[7] GONG Q, CHEN Y S. Parameter selection methods of delay and beamforming for cochlear implant speech enhancement[J]. Acoustic Physics, 2011, 57(4):542-550.
[8] MAJ J, WOUTERS J, MOONEN M. A two-stage adaptive beamformer for noise reduction in hearing aids[C]//Proceedings of 2001 Workshop on Acoustic Echo and Noise Control. Darmstadt, Germany:IEEE Press, 2001:171-174.
[9] LAI C C, NORDHOLM S, LEUNG Y H. Design of steerable spherical broadband beamformers with flexible sensor configurations[J]. IEEE Transactions on Audio, Speech and Language Processing, 2013, 21(2):427-438.
[10] LUO F L, YANG J, PAVLOVIC C. Adaptive null-forming scheme in digital hearing aids[J]. IEEE Transactions on Signal Processing, 2002, 50(7):1583-1590.
[11] JACEK D, JACOB B, SOFIENE A. Direction of arrival estimation using the parameterized spatial correlation matrix[J]. IEEE Transactions on Audio, Speech and Language Processing, 2007, 15(4):1327-1339.
[12] CHEN Y S, GONG Q. Broadband beamforming compensation algorithm in CI front-end acquisition[J/OL].[2017-10-01]. https://biomedical-engineering-online.biomedcentral.com/articles/10.1186/1475-925x-12-18.
[13] YOUSEFIAN N, LOIZOU P C. A dual-microphone speech enhancement algorithm based on the coherence function[J]. IEEE Transactions on Audio Speech and Language Processing, 2012, 20(2):599-609.
[14] AISSA-EL-BEY A, LINH-TRUNG N, ABED-MERAIM K, et al. Underdetermined blind separation of nondisjoint sources in the time-frequency domain[J]. IEEE Transactions on Signal Processing, 2007, 55(3):897-907.
[15] 崔杰, 肖灵, 王玥, 等. 一种用于数字助听器的WOLA滤波器组的设计准则[J]. 应用声学, 2010, 29(1):36-42.CUI J, XIAO L, WANG Y, et al. A kind of design criterion for WOLA filterbanks used in digital hearing aids[J]. Applied Acoustics, 2010, 29(1):36-42. (in Chinese)
[16] VARGA A, STEENEKEN H J M. Assessment for automatic speech recognition:Ⅱ. Noisex92:A database and an experiment to study the effect of additive noise on speech recognition systems[J]. Speech Communication, 1993, 12(3):247-251.
[1] 季建朝, 张宇, 王明新. 声传感器阵列风洞测量结果优化[J]. 清华大学学报(自然科学版), 2018, 58(1): 94-100.
Viewed
Full text


Abstract

Cited

  Shared   
  Discussed   
版权所有 © 《清华大学学报(自然科学版)》编辑部
本系统由北京玛格泰克科技发展有限公司设计开发 技术支持:support@magtech.com.cn