计算机科学 ›› 2015, Vol. 42 ›› Issue (10): 16-19.

• 目次 • 上一篇    下一篇

基于增益字典查询的语音增强算法

庞亮,陈 亮,张翼鹏,黄清泉   

  1. 解放军理工大学通信工程学院 南京210007,解放军理工大学通信工程学院 南京210007,解放军南京炮兵学院 南京211132,解放军理工大学通信工程学院 南京210007
  • 出版日期:2018-11-14 发布日期:2018-11-14

Speech Enhancement Based on Gain Dictionary Queries

PANG Liang, CHEN Liang, ZHANG Yi-peng and HUANG Qing-quan   

  • Online:2018-11-14 Published:2018-11-14

摘要: 对于基于统计模型的语音增强算法,不同分布模型对应于不同的增益函数,由于语音信号的不确定性,没有一种分布函数能准确对语音和噪声谱的分布建模,因此任何一种固定的统计模型均会存在一定的误差。所以提出一种增益字典查询的语音增强算法,该算法通过采用对数谱失真准则对一个语音噪声库进行增益的训练,得到一个增益的字典,其中输入为先验信噪比和后验信噪比的估计值。最后采用ITU-T P.826 PESQ、分段信噪比、总信噪比和对数谱失真对该算法进行了测试,并与基于高斯分布模型、拉普拉斯分布模型的算法进行了对比。实验结果表明,该算法无论在非平稳噪声还是平稳噪声环境下都比其他几种算法增强效果好,且音乐噪声和残留背景噪声也可以得到很好的抑制。

关键词: 语音增强,字典查询,判决引导,改进递归平均算法

Abstract: For speech enhancement algorithm based on statistical model,different distribution models are corresponding to different gain function,due to the uncertainty of the speech signal,no distribution function can accurately model the speech and noise spectra distribution,so any kind of fixed reference models will have some errors.We presented a gain dictionary queries based speech enhancement algorithm,getting a dictionary gain through training the voice of a noise library using log-spectral distortion criterion,for which the input is the estimate value of a priori and a posteriori SNR.Finally,we used ITU-T P.826 PESQ,segmented SNR,total SNR and log-spectral distortion criterion to test the proposed algorithm,and compared this algorithm with Gaussian distribution model and Laplace distribution model.The experimental results show that the algorithm is better than the other algorithms,whether in stationary or non-stationary noisy environments,and musical noise and residual background noise can be well suppressed.

Key words: Speech enhancement,Dictionary queries,Decision-directed,IMCRA

[1] Loizuo P C.Speech Enhancement Theory and Practice[M].CRC Press,2007:337-377
[2] 晏光华.一种基于MMSE-LSA和VAD的语音增强算法[J].移动通信,2014,0(4):59-62 Yan Guang-hua.A Speech Enhancement Algorithm Based on MMSE-LSA and VAD[J].Mobile Communication,2014,0(4):59-62
[3] 陈立伟,王文姝,袁頔.自适应高斯混合模型语音增强方法[J].应用科技,2009,6(7):11-15 Chen Li-wei,Wang Wen-shu,Yuan Di.A Speech Enhancement Method Based on Adaptive Gaussian Mixture Model[J].Applied Science and Technology,2009,36(7):11-15
[4] 梁岩,鲍长春,夏丙寅,等.基于高斯混合模型的压缩域语音增强方法[J].电子学报,2012,0(10):2031-2038 Liang Yan,Bao Chang-chun,Xia Bing-yin,et al.Compressed Domain Speech Enhancement Based on Gaussian Mixture Model[J].Acta Electronica Sinica,2012,0(10):2031-2038
[5] 邹霞,陈亮,张雄伟.基于Gamma语音模型的语音增强算法[J].通信学报,2006,7(10):118-123 Zou Xia,Chen Liang,Zhang Xiong-wei.Speech enhancement with Gamma speech modeling[J].Journal on Communications,2006,7(10):118-123
[6] 赵改华,周彬,张雄伟.修正的基于广义Gamma 语音模型语音增强算法[J].计算机工程与应用,2014,0(18):230-235 Zhao Gai-hua,Zhou Bin,Zhang Xiong-wei.Modified speech enhancement algorithm under signal presence probability with generalized Gamma speech model[J].Computer Engineering and Applications,2014,0(18):230-235
[7] 周彬,邹霞,张雄伟.基于多元Laplace语音模型的语音增强算法[J].电子与信息学报,2012,84(7):1562-1567 Zhou Bin,Zou Xia,Zhang Xiong-wei.Speech Enhancement with Multivariate Laplace Speech Model[J].Journal of Electronics & Information Technology,2012,84(7):1562-1567
[8] Wu D L,Zhu W P,Swamy M N S.Noise Spectrum Estimation with Improved Minimum Controlled Recursive Averaging based on Speech Enhancement Residue[C]∥IEEE International Midwest Symposium on Circuits and Systems(MWSCAS).Boise,USA,2012:945-951
[9] 杨波,王新房.基于非因果先验信噪比估计的语音增强改进算法[J].计算机系统应用,2012,1(7):200-202 Yang Bo,Wang Xin-fang.Improved Speech Enhancement Algorithm Based on Noncausal a Priori SNR Estimation[J].Compu-ter Systems & Applications,2012,1(7):200-202
[10] Yong P C,Nordholm S,Dam H H.Trade-off Evaluation forSpeech Enhancement Algorithms with Respect to The a Priori SNR Estimation[C]∥IEEE International Conference on Acoustics,Speech and Signal Processing(ICASSP).Kyoto,Japan,2012:4657-4660
[11] Ekelens J,Jensen J,Heusdens R.A Data-Driven Approach toOptimizing Spectral Speech Enhancement Methods for Various Error Criteria[J].Speech Communication,2007,49(5):530-541

No related articles found!
Viewed
Full text


Abstract

Cited

  Shared   
  Discussed   
No Suggested Reading articles found!