Low SNR denoising algorithm based on adaptive voice activity detection and minimum mean-square error log-spectral amplitude estimation

doi:10.11772/j.issn.1001-9081.2019111880

Journal of Computer Applications ›› 2020, Vol. 40 ›› Issue (6): 1763-1768.DOI: 10.11772/j.issn.1001-9081.2019111880

• Network and communications • Previous Articles Next Articles

Low SNR denoising algorithm based on adaptive voice activity detection and minimum mean-square error log-spectral amplitude estimation

ZHANG Haoran, WANG Xueyuan, LI Xiaoxia

Sichuan Key Laboratory of Special Environmental Robotics (School of Information Engineering, Southwest University of Science and Technology), Mianyang Sichuan 621010, China

Received:2019-11-05 Revised:2020-01-03 Online:2020-06-10 Published:2020-06-18
Contact: WANG Xueyuan, born in 1974, Ph. D., associate professor. His research interests include information processing.
About author:ZHANG Haoran, born in 1995, M. S. candidate. His research interests include signal processing, speech enhancement.WANG Xueyuan, born in 1974, Ph. D., associate professor. His research interests include information processing.LI Xiaoxia, born in 1976, Ph. D., professor. Her research interests include pattern recognition, computer vision.
Supported by:
National Natural Science Foundation of China(61771411).

基于自适应阈值活动语音检测和最小均方误差对数谱幅度估计的低信噪比降噪算法

张皓然, 王学渊, 李小霞

特殊环境机器人技术四川省重点实验室(西南科技大学信息工程学院)，四川绵阳 621010

通讯作者: 王学渊(1974—)
作者简介:张皓然(1995—)，男，四川绵阳人，硕士研究生，主要研究方向：信号处理、语音增强.王学渊(1974—)，男，四川绵阳人，副教授，博士，主要研究方向：信息处理.李小霞(1976—)，女，四川安岳人，教授，博士，主要研究方向：模式识别、计算机视觉.
基金资助:
国家自然科学基金资助项目(61771411)。

Abstract

Abstract: Aiming at the limitations of traditional noise reduction methods for acoustic signals in low Signal-to-Noise Ratio (SNR) environment, a real-time noise reduction algorithm was proposed by combining adaptive threshold Voice Activity Detection (VAD) algorithm and Minimum Mean-Square Error Log-Spectral Amplitude estimation (MMSE-LSA). Firstly, the background noise was estimated in VAD algorithm by probability statistics based on the maximum value of the energy probability, and the obtained background noise was updated in real time and saved. Then, the background noise updated in real time was used as the reference noise of MMSE-LSA, and the noise amplitude spectrum was updated adaptively. Finally, the noise reduction processing was performed. The experimental results on four kinds of acoustic signals in real scenes show that the proposed algorithm can guarantee the real-time processing of low SNR acoustic signals; and compared with the traditional MMSE-LSA algorithm, it has the SNR of the noise reduction signal increased by 10-15 dB without over-subtraction. It can be applied to practical engineering.

Key words: real environment, adaptive threshold, Voice Activity Detection (VAD) algorithm, real-time Minimum Mean-Square Error Log-Spectral Amplitude estimation (MMSE-LSA) algorithm, real-time background, low Signal-to-Noise Ratio (SNR)

摘要： 针对低信噪比(SNR)环境下传统方法对声信号降噪的局限性，提出了一种联合自适应阈值活动语音检测（VAD）算法和最小均方误差对数谱幅度估计（MMSE-LSA）的实时降噪算法。首先，在VAD算法中通过基于能量概率最大值的概率统计来对背景噪声进行估计，对得到的背景噪声进行实时更新并保存；然后，将实时更新的背景噪声作为MMSE-LSA的参考噪声，并对噪声幅度谱进行自适应更新，最后进行降噪处理。通过在真实场景中对四类声信号进行实验，结果表明，该算法在保证对低SNR声信号的实时处理的情况下，相较于传统MMSE-LSA算法，降噪信号的SNR能够提高10~15 dB，且不存在信号过减的情况，可应用于实际工程。

关键词: 真实环境, 自适应阈值, 活动语音检测算法, 实时最小均方误差对数谱幅度估计算法, 实时背景, 低信噪比

CLC Number:

TP912.35

ZHANG Haoran, WANG Xueyuan, LI Xiaoxia. Low SNR denoising algorithm based on adaptive voice activity detection and minimum mean-square error log-spectral amplitude estimation[J]. Journal of Computer Applications, 2020, 40(6): 1763-1768.

张皓然, 王学渊, 李小霞. 基于自适应阈值活动语音检测和最小均方误差对数谱幅度估计的低信噪比降噪算法[J]. 计算机应用, 2020, 40(6): 1763-1768.

References

1 UDREAR M, CIOCHINAR M, VIZIREANUD N. Reduction of background noise from affected speech using a spectral subtraction algorithm based on masking properties of the human ear[C]// Proceedings of the 7th International Conference on Telecommunication in Modern Satellite, Cable and Broadcasting Services. Piscataway: IEEE, 2005:135-138.
2 SIDDAPPAJI, SUDHAK L. Performance analysis of New Time Varying LMS (NTVLMS) adaptive filtering algorithm in noise cancellation system[C]// Proceedings of the 2015 International Conference on Communication, Information and Computing Technology. Piscataway: IEEE, 2015:1-6.
3 MODHAVEN, KARUNAY, TONDES. Design of matrix Wiener filter for noise reduction and speech enhancement in hearing aids[C]// Proceedings of the 2016 IEEE International Conference on Recent Trends in Electronics, Information and Communication Technology. Piscataway: IEEE, 2016: 843-847.
4 WEIJ, WANGM, YANGH. MMSE-LSA based wavelet threshold denoising algorithm for low SNR speech [C]// Proceedings of the 2016 IEEE International Conference on Digital Signal Processing. Piscataway: IEEE, 2016: 253-256.
5 郭武,马啸空.复杂噪声场景下的活动语音检测方法[J].清华大学学报(自然科学版),2016,56(11):1190-1195. GUOW, MAX K. Voice activity detection in complex noise environment [J]. Journal of Tsinghua University (Science and Technology), 2016, 56(11): 1190-1195.
6 李文洁,张晴晴,张鹏远,等.基于维特比算法的深度神经网络语音端点检测[J].重庆邮电大学学报(自然科学版),2018,30(2):210-215. LIW J, ZHANGQ Q, ZHANGP Y, et al. Voice activity detection based on deep neural networks with Viterbi [J]. Journal of Chongqing University of Posts and Telecommunications (Natural Science Edition), 2018, 30(2): 210-215.
7 FUKANEA R, SAHARES L. Noise estimation algorithms for speech enhancement in highly non-stationary environments [J]. International Journal of Computer Science Issues, 2011, 8(2): 39-44.
8 EPHRAIMY, MALAHD. Speech enhancement using a minimum mean-square error log-spectral amplitude estimator [J]. IEEE Transactions on Acoustics, Speech, and Signal Processing, 1985, 33(2): 443-445.)
9 LIX, LIG, LIX. Improved voice activity detection based on iterative spectral subtraction and double thresholds for CVR [C]// Proceedings of the 2008 Workshop on Power Electronics and Intelligent Transportation System. Piscataway: IEEE, 2008: 153-156.
10 BAHRAMIM, SEYEDINS. MMSE log-spectral amplitude estimation for single channel speech enhancement under speech presence uncertainty by Weibull speech priors [C]// Proceedings of the 2018 Iranian Conference on Electrical Engineering. Piscataway: IEEE, 2018: 749-754.
11 ZHANGY, LIUY. An improved MMSE-LSA speech enhancement algorithm based on human auditory masking property [C]// Proceedings of the 2013 International Conference on Asian Language Processing. Piscataway: IEEE, 2013: 151-154.
12 YKHLEFF, YKHLEFH. A smoothed minimum mean-square error log-spectral amplitude estimator for speech enhancement [C]// Proceedings of the 2014 International Conference on Multimedia Computing and Systems. Piscataway: IEEE, 2014:246-249.
13 冯炎.基于直接判决估计和预测估计的语音增强算法[J].信息与电子工程,2010,8(1):76-78. FENGY. Speech enhancement based on Decision-Directed incorporating predicted estimation [J]. Information and Electronic Engineering, 2010, 8(1): 76-78.
14 HOUX, GUOS, CUIH, et al. Speech enhancement for non-stationary noise environments [C]// Proceedings of the 2009 International Conference on Information Engineering and Computer Science. Piscataway: IEEE, 2009:1-3.
15 严思伟,屈晓旭,娄景艺.基于连续噪声谱估计的谱减法语音增强算法[J].通信技术,2018,51(6):1296-1301. YANS W, QUX X, LOUJ Y. Spectral subtraction speech enhancement algorithm based on continuous noise spectrum estimation [J]. Communications Technology, 2018, 51(6): 1296-1301.
16 李英杰.基于声谱图的音频事件检测特征提取研究[D].北京:北京邮电大学,2017:7-8. LIY J. Research on feature extraction of audio event detection based on spectrogram [D].Beijing: Beijing University of Posts and Telecommunications,2017:7-8.
17 ROMPAEY RVAN, MOONENM. GEVD based speech and noise correlation matrix estimation for multichannel wiener filter based noise reduction [C]// Proceedings of the 26th European Signal Processing Conference. Piscataway: IEEE, 2018: 2544-2548.
18 FODORB, FINGSCHEIDTT. MMSE log-spectral amplitude estimation under speech presence uncertainty using generalized gamma speech priors [C]// Proceedings of the 2012 International Workshop on Acoustic Signal Enhancement. Frankfurt: VDE, 2012:1-4.
19 何伟俊,贺前华,吴俊峰,等.基于子带双特征的自适应保留似然比鲁棒语音检测算法[J].电子与信息学报,2016,38(11):2879-2886. HEW J, HEQ H, WUJ F, et al. Adaptively reserved likelihood ratio-based robust voice activity detection with sub-band double features [J]. Journal of Electronics and Information Technology, 2016, 38(11): 2879-2886.

[1]	WANG Sipeng, DU Changping, YE Zhixian, SONG Guanghua, ZHENG Yao. Soft fault detection for flapping wing micro aerial vehicle based on multistep neural network observer [J]. Journal of Computer Applications, 2020, 40(8): 2449-2454.
[2]	WANG Zhongfeng, XU Zhiyuan, SONG Chunhe, ZHANG Hongyu, CAI Yingkai. Gradient-based deep network pruning algorithm [J]. Journal of Computer Applications, 2020, 40(5): 1253-1259.
[3]	WANG Zhizhong, LI Hongyi, HAN Chuang. QRS complex detection algorithm of electrocardiograph based on Shannon energy and adaptive threshold [J]. Journal of Computer Applications, 2020, 40(1): 304-310.
[4]	DING Lianjing, LIU Guangshuai, LI Xurui, CHEN Xiaowen. Face recognition combining weighted information entropy with enhanced local binary pattern [J]. Journal of Computer Applications, 2019, 39(8): 2210-2216.
[5]	XU Dong, LI Yong, LIU Dongdong, LU Yakai. Weight adjustable interference alignment algorithm with imperfect channel state information [J]. Journal of Computer Applications, 2018, 38(8): 2370-2374.
[6]	PENG Zheng, CHEN Dongfang, WANG Xiaofeng. Adaptive threshold denoising of regularized super-resolution reconstruction procedure [J]. Journal of Computer Applications, 2017, 37(7): 2084-2088.
[7]	HUANG Yonghang, TANG Yong, LI Chunying, TANG Zhikang, LIU Jiwei. Academic paper recommendation model based on community partition [J]. Journal of Computer Applications, 2016, 36(5): 1279-1283.
[8]	LAN Li, HE Xiaohai, WU Xiaohong, TENG Qizhi. Adaptive shadow removal based on superpixel and local color constancy [J]. Journal of Computer Applications, 2016, 36(10): 2837-2841.
[9]	LYU Duliang, JIA Zhenhong, YANG Jie, Nikola KASABOV. Remote sensing image enhancement based on combination of non-subsampled shearlet transform and guided filtering [J]. Journal of Computer Applications, 2016, 36(10): 2880-2884.
[10]	LI Chunying, TANG Yong, TANG Zhikang, HUANG Yonghang, YUAN Chengzhe, ZHAO Jiandong. Community detection model in large scale academic social networks [J]. Journal of Computer Applications, 2015, 35(9): 2565-2568.
[11]	LYU Jiaqing, LIU Licheng, HAO Luguo, ZHANG Wenzhong. Adaptive moving object extraction algorithm based on visual background extractor [J]. Journal of Computer Applications, 2015, 35(7): 2029-2032.
[12]	JI Xiaofei, QIN Ningli. Design and implementation of multi-target detection and recognition algorithm for optical remote sensing image [J]. Journal of Computer Applications, 2015, 35(11): 3302-3307.
[13]	YANG Zhi LUO Guo YUAN Fangfang. Application of adaptive wavelet scalogram threshold in diaphragmatic electromyographic signal denoising [J]. Journal of Computer Applications, 2013, 33(09): 2679-2682.
[14]	WANG Junqing ZHANG Weichuan WANG Fuping CHEN Meirong. Corner detection algorithm using multi-chord curvature polynomial [J]. Journal of Computer Applications, 2013, 33(08): 2313-2316.
[15]	ZHANG Fan PENG Zhong-wei MENG Shui-jin. Improved Canny edge detection method based on self-adaptive threshold [J]. Journal of Computer Applications, 2012, 32(08): 2296-2298.

Low SNR denoising algorithm based on adaptive voice activity detection and minimum mean-square error log-spectral amplitude estimation

基于自适应阈值活动语音检测和最小均方误差对数谱幅度估计的低信噪比降噪算法

PDF

Knowledge

Abstract

Cite this article

share this article

References

Related Articles 15

Recommended Articles

Metrics