高级搜索

留言板

尊敬的读者、作者、审稿人, 关于本刊的投稿、审稿、编辑和出版的任何问题, 您可以本页添加留言。我们将尽快给您答复。谢谢您的支持!

姓名
邮箱
手机号码
标题
留言内容
验证码

色噪声下基于白化频谱重排鲁棒主成分分析的语音增强算法

罗勇江 杨腾飞 赵冬

罗勇江, 杨腾飞, 赵冬. 色噪声下基于白化频谱重排鲁棒主成分分析的语音增强算法[J]. 电子与信息学报, 2021, 43(12): 3671-3679. doi: 10.11999/JEIT200594
引用本文: 罗勇江, 杨腾飞, 赵冬. 色噪声下基于白化频谱重排鲁棒主成分分析的语音增强算法[J]. 电子与信息学报, 2021, 43(12): 3671-3679. doi: 10.11999/JEIT200594
Yongjiang LUO, Tengfei YANG, Dong ZHAO. Speech Enhancement Algorithm Based on Robust Principal Component Analysis with Whitened Spectrogram Rearrangement in Colored Noise[J]. Journal of Electronics & Information Technology, 2021, 43(12): 3671-3679. doi: 10.11999/JEIT200594
Citation: Yongjiang LUO, Tengfei YANG, Dong ZHAO. Speech Enhancement Algorithm Based on Robust Principal Component Analysis with Whitened Spectrogram Rearrangement in Colored Noise[J]. Journal of Electronics & Information Technology, 2021, 43(12): 3671-3679. doi: 10.11999/JEIT200594

色噪声下基于白化频谱重排鲁棒主成分分析的语音增强算法

doi: 10.11999/JEIT200594
详细信息
    作者简介:

    罗勇江:男,1979年生,副教授,研究方向为低秩稀疏分解、高速信号处理

    杨腾飞:男,1993年生,硕士,研究方向为语音增强、语音信号处理

    赵冬:男,1996年生,硕士生,研究方向为非高斯信号处理

    通讯作者:

    罗勇江 yjluo@mail.xidian.edu.cn

  • 中图分类号: TN912.35

Speech Enhancement Algorithm Based on Robust Principal Component Analysis with Whitened Spectrogram Rearrangement in Colored Noise

  • 摘要: 基于鲁棒主成分分析(RPCA)的单通道语音增强算法是高斯白噪声环境下语音增强的一种重要处理手段,但其对低秩语音分量处理效果欠佳且无法较好地抑制色噪声。针对此问题,该文提出一种基于白化频谱重排RPCA的改进语音增强算法(WSRRPCA),通过优化噪声白化模型,将色噪声语音增强转换成白噪声语音信号处理,利用频谱重排改进RPCA语音增强处理算法,从而获得色噪声环境下语音信号处理性能的整体提升。仿真实验表明,该算法能够较好地实现色噪声环境下的语音增强,且相对于其他算法具有更佳的噪声抑制和语音质量提升能力。
  • 图  1  基于RPCA的单通道语音增强算法的系统框图

    图  2  WSRRPCA算法的结构框图

    图  3  不同信噪比下SDR与$\lambda $的关系曲线

    图  4  WSRRPCA与RPCA处理结果对比图

    图  5  buccaneer1, buccaneer2和f16噪声环境下不同算法的性能对比图

    图  6  Factory1, hfchannel和pink噪声环境下不同算法的性能对比图

    表  1  不同噪声下多种算法的性能对比

    噪声类型语音增强算法SDR (dB)PESQ
    GASS1.11481.5057
    logMMSE-SPU–2.95660.9222
    buccaneer1RPCA4.84321.6275
    CLSMD5.35831.0624
    WSRRPCA6.25301.6106
    GASS–0.49801.5690
    logMMSE-SPU–3.02101.1192
    buccaneer2RPCA3.84811.7261
    CLSMD4.61470.9079
    WSRRPCA4.99891.6944
    GASS1.48051.7816
    logMMSE-SPU–2.32101.1926
    f16RPCA4.38861.8461
    CLSMD5.46811.1948
    WSRRPCA6.20301.8751
    GASS0.31331.4930
    logMMSE-SPU–2.76921.1512
    factory1RPCA4.08861.8264
    CLSMD4.26911.2895
    WSRRPCA5.11381.7905
    GASS1.41681.3519
    logMMSE-SPU–3.01501.0336
    hfchannelRPCA5.17691.6378
    CLSMD6.77711.1689
    WSRRPCA6.14181.6441
    GASS1.00081.6570
    logMMSE-SPU–1.40771.2425
    pinkRPCA4.08351.8472
    CLSMD3.98051.4122
    WSRRPCA7.06991.9045
    下载: 导出CSV
  • [1] LOIZOU P C. Speech Enhancement: Theory and Practice[M]. 2nd ed. London: CRC Press, 2013: 1–2.
    [2] BOLL S. Suppression of acoustic noise in speech using spectral subtraction[J]. IEEE Transactions on Acoustics, Speech, and Signal Processing, 1979, 27(2): 113–120. doi: 10.1109/tassp.1979.1163209
    [3] EPHRAIM Y and MALAH D. 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. doi: 10.1109/TASSP.1985.1164550
    [4] SCALART P and FILHO J V. Speech enhancement based on a priori signal to noise estimation[C]. International Conference on Acoustics, Speech, and Signal Processing, Atlanta, 1996: 629–632. doi: 10.1109/ICASSP.1996.543199.
    [5] EPHRAIM Y and VAN TREES G L. A signal subspace approach for speech enhancement[J]. IEEE Transactions on Speech and Audio Processing, 1995, 3(4): 251–266. doi: 10.1109/89.397090
    [6] YI Hu and LOIZOU P C. A generalized subspace approach for enhancing speech corrupted by colored noise[J]. IEEE Transactions on Speech and Audio Processing, 2003, 11(4): 334–341. doi: 10.1109/TSA.2003.814458
    [7] SUN Chengli, ZHANG Qin, WANG Jian, et al. Noise reduction based on robust principal component analysis[J]. Journal of Computational Information Systems, 2014, 10(10): 4403–4410. doi: 10.12733/jcis10408
    [8] HUANG Jianjun, ZHANG Xiongwei, ZHANG Yafei, et al. Speech denoising via low-rank and sparse matrix decomposition[J]. ETRI Journal, 2014, 36(1): 167–170. doi: 10.4218/etrij.14.0213.0033
    [9] MAVADDATY S, AHADI S M, and SEYEDIN S. A novel speech enhancement method by learnable sparse and low-rank decomposition and domain adaptation[J]. Speech Communication, 2016, 76: 42–60. doi: 10.1016/j.specom.2015.11.003
    [10] SUN Pengfei and QIN Jun. Low-rank and sparsity analysis applied to speech enhancement via online estimated dictionary[J]. IEEE Signal Processing Letters, 2016, 23(12): 1862–1866. doi: 10.1109/lsp.2016.2627029
    [11] LUO Yongjiang and MAO Yu. Single-channel speech enhancement based on multi-band spectrogram-rearranged RPCA[J]. Electronics Letters, 2019, 55(7): 415–417. doi: 10.1049/el.2018.8131
    [12] CANDÈS E J, LI Xiaodong, MA Yi, et al. Robust principal component analysis?[J]. Journal of the ACM, 2011, 58(3): 11. doi: 10.1145/1970392.1970395
    [13] NAZIH M, MINAOUI K, and COMON P. Using the proximal gradient and the accelerated proximal gradient as a canonical polyadic tensor decomposition algorithms in difficult situations[J]. Signal Processing, 2020, 171: 107472. doi: 10.1016/j.sigpro.2020.107472
    [14] FENG Peihua, LING B W K, LEI Ruisheng, et al. Singular spectral analysis-based denoising without computing singular values via augmented Lagrange multiplier algorithm[J]. IET Signal Processing, 2019, 13(2): 149–156. doi: 10.1049/iet-spr.2018.5086
    [15] LEI Yunwen and ZHOU Dingxuan. Analysis of singular value thresholding algorithm for matrix completion[J]. Journal of Fourier Analysis and Applications, 2019, 25(6): 2957–2972. doi: 10.1007/s00041-019-09688-8
    [16] JARAMILLO A E, NIELSEN J K, CHRISTENSEN M G, et al. A study on how pre-whitening influences fundamental frequency estimation[C]. International Conference on Acoustics, Speech and Signal Processing, Brighton, England, 2019: 6495–6499. doi: 10.1109/ICASSP.2019.8683653.
    [17] VASEGHI S V. Advanced Digital Signal Processing and Noise Reduction[M]. 4th ed. Hoboken: John Wiley & Sons, 2008: 229–230.
    [18] SMITH III J O. Spectral Audio Signal Processing[M]. W3K Publishing, USA, 2011: 298–301.
    [19] 张明, 刘祥楼, 姜峥嵘. 基于LPC的语音信号预测仿真分析[J]. 光学仪器, 2015, 37(1): 71–74. doi: 10.3969/j.issn.1005-5630.2015.01.015

    ZHANG Ming, LIU Xianglou, and JIANG Zhengrong. Simulation analysis of speech signal prediction based on LPC[J]. Optical Instruments, 2015, 37(1): 71–74. doi: 10.3969/j.issn.1005-5630.2015.01.015
    [20] KAPRALOV M. Sparse Fourier transform in any constant dimension with nearly-optimal sample complexity in sublinear time[C]. The Forty-eighth Annual ACM Symposium on Theory of Computing, Virtual Event, Italy, 2016: 264–277. doi: 10.1145/2897518.2897650.
    [21] WANG D L and LIM J S. The unimportance of phase in speech enhancement[J]. IEEE Transactions on Acoustics, Speech, and Signal Processing, 1982, 30(4): 679–681. doi: 10.1109/TASSP.1982.1163920
    [22] VINCENT E, GRIBONVAL R, and FEVOTTE C. Performance measurement in blind audio source separation[J]. IEEE Transactions on Audio, Speech, and Language Processing, 2006, 14(4): 1462–1469. doi: 10.1109/TSA.2005.858005
    [23] RAM R and MOHANTY M N. Use of radial basis function network with discrete wavelet transform for speech enhancement[J]. International Journal of Computational Vision and Robotics, 2019, 9(2): 207–223. doi: 10.1504/IJCVR.2019.10019996
    [24] SUN Chengli, ZHU Qi, and WAN Minghua. A novel speech enhancement method based on constrained low-rank and sparse matrix decomposition[J]. Speech Communication, 2014, 60: 44–55. doi: 10.1016/j.specom.2014.03.002
    [25] LU Yang and LOIZOU P C. A geometric approach to spectral subtraction[J]. Speech Communication, 2008, 50(6): 453–466. doi: 10.1016/j.specom.2008.01.003
    [26] COHEN I. Optimal speech enhancement under signal presence uncertainty using log-spectral amplitude estimator[J]. IEEE Signal Processing Letters, 2002, 9(4): 113–116. doi: 10.1109/97.1001645
  • 加载中
图(6) / 表(1)
计量
  • 文章访问数:  639
  • HTML全文浏览量:  427
  • PDF下载量:  74
  • 被引次数: 0
出版历程
  • 收稿日期:  2020-07-20
  • 修回日期:  2021-03-25
  • 网络出版日期:  2021-06-03
  • 刊出日期:  2021-12-21

目录

    /

    返回文章
    返回