基于小波分析与Boll改进谱减法的普米语语音增强算法
|
摘要
小波分析和Boll改进谱减法具有很好的减噪效果,但同时也存在"音乐噪声"和语音失真.为了解决这2个问题,基于这2种算法,研究了一种它们的联合算法并对加噪后的普米语进行语音减噪实验.通过对减噪后语音的试听和波形图的观察以及信噪比的对比,对实验的结果进行分析.实验结果表明,针对叠加了随机高斯白噪声的普米语,算法能有效地减少这2种方法单独使用时所产生的"音乐噪声"和语音失真.
Wavelet analysis and improved Boll's spectral subtraction have good noise reduction effects, but there are also "music noise" and speech distortion. In order to solve the two problems, this paper studies a joint algorithm based on these two algorithms and carries out speech denoising experiments on the Pumi language after noise addition. The experimental results are analyzed by comparing the audition and waveform of the noise-reduced speech and the comparison of the signal-to-noise ratio. The experimental results show that for the Pumi language with random Gaussian white noise superimposed, this joint algorithm can effectively reduce the "music noise" and speech distortion produced by the two methods when they are used separately.
Wavelet analysis and improved Boll's spectral subtraction have good noise reduction effects, but there are also "music noise" and speech distortion. In order to solve the two problems, this paper studies a joint algorithm based on these two algorithms and carries out speech denoising experiments on the Pumi language after noise addition. The experimental results are analyzed by comparing the audition and waveform of the noise-reduced speech and the comparison of the signal-to-noise ratio. The experimental results show that for the Pumi language with random Gaussian white noise superimposed, this joint algorithm can effectively reduce the "music noise" and speech distortion produced by the two methods when they are used separately.
引文
[1] 张俊.基于谱减算法语音增强的研究[J].信息技术,2009(3):74-76.
[2] 肖全宝,徐晨,宋广为,等.用于语音识别的基于高谱分辨率的谱减法[J].广西师范大学学报(自然科学版),2006,24(4):26-29.
[3] 崔周培,戴冬原.应用谱减法处理带噪语音信号的研究[J].仪表技术,2011(6):46-49.
[4] 吴亚琴.基于Matlab的谱减法语音增强的研究[J].电子世界,2014(3):119-119.
[5] BOLL S.Suppression of acoustic noise in speech using spectral subtraction[J].Acoustics Speech & Signal Processing IEEE Transactions on,1979,27(2):113-120.
[6] EPHRAIM Y,TREES H L V .A signal subspace approach for speech enhancement[J].IEEE Transactions on Speech and Audio Processing,2002,3(4):251-266.
[7] HU Y H Y,LOIZOU P C .Speech enhancement based on wavelet thresholding the multitaper spectrum[J].IEEE Transactions on Speech & Audio Processing,2004,12(1):59-67.
[8] 李潇,李宏.一种改进的基于DCT变换的语音增强算法[J].计算机仿真,2010,27(12):376-380.
[9] A.Z M S .DCT speech enhancement with hard and soft thresholds criteria[J].Electronics Letters,2002,38(13):669-670.
[10] 楼红伟,胡光锐.基于简化的KLT和小波变换的非平稳宽带噪声语音增强[J].控制与决策,2003,18(5):577-580.
[11] LIU W,ZHENG N.Non-negative matrix factorization based methods for object recognition[J].Pattern Recognition Letters,2004,25(8):893-897.
[12] LUSTIG M,DONOHO D L,SANTOS J M.Compressed sensing MRI[J].IEEE Signal Processing Magazine,2008,25(2):72-82.
[13] 张飞.一种基于小波变换的语音信号去噪净化方法[J].电子测量技术,2009,32(6):68-71.
[14] 姚远,王秋菊,周伟,等.改进谱减法结合神经网络的语音增强研究[J].电子测量技术,2017,40(7):75-79.
[15] 谢斌,乐鸿浩,陈博.基于小波去噪与离散余弦变换相结合的正交频分复用系统信道估计算法[J].计算机应用,2015,35(9):2461-2464.
[16] 马明.基于谱减法和小波阈值分解的联合消噪算法研究[J].无线互联科技,2011(3):6-9.
[17] 张洪涛,裴浩,周航.谱减法与小波变换联合语音增强研究[J].湖北工业大学学报,2013,28(1):64-66.
[18] 朱建华,田岚.基于谱减法和小波阈值的新型语音降噪算法研究[J].煤炭技术,2013,32(1):171-173.
[19] 李余芳,苏洁,胡文君,等.基于 HTK 的普米语孤立词的语音识别[J].云南民族大学学报 (自然科学版),2015,24(5):426-430.
[20] 杨花,江涛,董华珍,等.基于 PSO-SVM 的普米语语谱图识别[J].云南民族大学学报 (自然科学版),2018,27(05):77-81.
[21] 董华珍,潘文林,王翠,等.基于免疫遗传优化支持向量机的普米语孤立词语谱图分类[J].云南民族大学学报 (自然科学版),2019,28 (1):100-104.
[2] 肖全宝,徐晨,宋广为,等.用于语音识别的基于高谱分辨率的谱减法[J].广西师范大学学报(自然科学版),2006,24(4):26-29.
[3] 崔周培,戴冬原.应用谱减法处理带噪语音信号的研究[J].仪表技术,2011(6):46-49.
[4] 吴亚琴.基于Matlab的谱减法语音增强的研究[J].电子世界,2014(3):119-119.
[5] BOLL S.Suppression of acoustic noise in speech using spectral subtraction[J].Acoustics Speech & Signal Processing IEEE Transactions on,1979,27(2):113-120.
[6] EPHRAIM Y,TREES H L V .A signal subspace approach for speech enhancement[J].IEEE Transactions on Speech and Audio Processing,2002,3(4):251-266.
[7] HU Y H Y,LOIZOU P C .Speech enhancement based on wavelet thresholding the multitaper spectrum[J].IEEE Transactions on Speech & Audio Processing,2004,12(1):59-67.
[8] 李潇,李宏.一种改进的基于DCT变换的语音增强算法[J].计算机仿真,2010,27(12):376-380.
[9] A.Z M S .DCT speech enhancement with hard and soft thresholds criteria[J].Electronics Letters,2002,38(13):669-670.
[10] 楼红伟,胡光锐.基于简化的KLT和小波变换的非平稳宽带噪声语音增强[J].控制与决策,2003,18(5):577-580.
[11] LIU W,ZHENG N.Non-negative matrix factorization based methods for object recognition[J].Pattern Recognition Letters,2004,25(8):893-897.
[12] LUSTIG M,DONOHO D L,SANTOS J M.Compressed sensing MRI[J].IEEE Signal Processing Magazine,2008,25(2):72-82.
[13] 张飞.一种基于小波变换的语音信号去噪净化方法[J].电子测量技术,2009,32(6):68-71.
[14] 姚远,王秋菊,周伟,等.改进谱减法结合神经网络的语音增强研究[J].电子测量技术,2017,40(7):75-79.
[15] 谢斌,乐鸿浩,陈博.基于小波去噪与离散余弦变换相结合的正交频分复用系统信道估计算法[J].计算机应用,2015,35(9):2461-2464.
[16] 马明.基于谱减法和小波阈值分解的联合消噪算法研究[J].无线互联科技,2011(3):6-9.
[17] 张洪涛,裴浩,周航.谱减法与小波变换联合语音增强研究[J].湖北工业大学学报,2013,28(1):64-66.
[18] 朱建华,田岚.基于谱减法和小波阈值的新型语音降噪算法研究[J].煤炭技术,2013,32(1):171-173.
[19] 李余芳,苏洁,胡文君,等.基于 HTK 的普米语孤立词的语音识别[J].云南民族大学学报 (自然科学版),2015,24(5):426-430.
[20] 杨花,江涛,董华珍,等.基于 PSO-SVM 的普米语语谱图识别[J].云南民族大学学报 (自然科学版),2018,27(05):77-81.
[21] 董华珍,潘文林,王翠,等.基于免疫遗传优化支持向量机的普米语孤立词语谱图分类[J].云南民族大学学报 (自然科学版),2019,28 (1):100-104.