基于先驗(yàn)信噪比的改進(jìn)型譜減法.doc
約31頁DOC格式手機(jī)打開展開
基于先驗(yàn)信噪比的改進(jìn)型譜減法,摘 要語音增強(qiáng)的目的主要是改進(jìn)語音質(zhì)量,在消除背景噪音的同時提高語音可懂度,但是這兩個目的往往不能兼得。目前有一些對非平穩(wěn)噪聲干擾下的語音信號進(jìn)行增強(qiáng)的方法,可以降低背景噪聲,但有時會引入刺耳的音樂噪聲,且不能提高語音的可懂度,甚至略有下降。其中譜減法作為一種單信道語音增強(qiáng)方法,以其簡單有效...
![](http://img.queshao.com/images/pcgzh.gif)
![](http://preview.queshao.com/tobuy/208109.gif)
內(nèi)容介紹
此文檔由會員 wanli1988go 發(fā)布
基于先驗(yàn)信噪比的改進(jìn)型譜減法
摘 要
語音增強(qiáng)的目的主要是改進(jìn)語音質(zhì)量,在消除背景噪音的同時提高語音可懂度,但是這兩個目的往往不能兼得。目前有一些對非平穩(wěn)噪聲干擾下的語音信號進(jìn)行增強(qiáng)的方法,可以降低背景噪聲,但有時會引入刺耳的音樂噪聲,且不能提高語音的可懂度,甚至略有下降。其中譜減法作為一種單信道語音增強(qiáng)方法,以其簡單有效在使用化的語音增強(qiáng)應(yīng)用中深受歡迎。
本文研究采用閥值法法對非平穩(wěn)背景噪聲信號進(jìn)行估計(jì),計(jì)算出先驗(yàn)信噪比,得到還原的純凈語音信號。本文用MATLAB實(shí)現(xiàn)了整個算法的仿真,并與傳統(tǒng)譜減法結(jié)果相比較,仿真結(jié)果表明,該算法對非平穩(wěn)噪聲追蹤性較好,在抑制背景噪聲,減少音樂噪聲前提下,提高了語音的可懂度,其計(jì)算復(fù)雜度也可以接受。
關(guān)鍵詞: 譜減法 語音增強(qiáng) 噪聲估計(jì)
Abstract
The mainly purpose of speech enhancement is to improve the quality of speech, raises the intelligibility of speech while eliminating background noise, but this two purposes can not often be gotten simultaneously. Now there are many method of speech enhancement for the speech signal with near-stationary noise can reduce background noise, but can produce an annoying noise called music noise, and the intelligibility can not raise even drops. As a method of one channel speech enhancement, spectral subtraction is deeply welcome because it is easy and efficient.
In this paper, the SNR of the speech signal is estimated, then the enhanced speech signal is gotten. MATLAB is used to realize the purpose, and it is compared with traditional spectral subtraction. It shows that this algorithm is better for the speech signal with the near-stationary noise; it restrains the music noise and reduces the background noise, it also raises the intelligibility. The complex degree also can be accepted.
Keyword: Spectral subtraction Speech enhancement
Noise estimation
目錄
第一章 緒論 1
1.1 課題背景及問題提出 1
1.2 語音增強(qiáng)研究的主要內(nèi)容和發(fā)展方向 1
1.3 譜減法語音增強(qiáng)的研究現(xiàn)狀 3
第二章 語音感知和噪聲特性 5
2.1 語音特性 5
2.2噪聲特性 5
第三章 譜減法語音增強(qiáng)算法 7
3.1 語音增強(qiáng)算法概述 7
3.2一般譜減法 8
3.2.1 基本原理 8
3.2.3 基本步驟和仿真結(jié)果 11
3. 3 “音樂噪聲”的產(chǎn)生 13
3.4 改進(jìn)的譜減法 14
3.4.1改進(jìn)型譜減法的種類 14
3.4.2 幅度譜平滑 15
3.4.3 被減項(xiàng)加權(quán)值處理 16
3.4.4 功率譜修正處理 16
第四章 基于先驗(yàn)信噪比估計(jì)的改進(jìn)型譜減法 19
4.1 基本原理 19
4.2 寂靜幀和語音幀的判定 19
4.3 算法的實(shí)現(xiàn) 21
4.4 參數(shù)的估計(jì) 22
4.5 實(shí)驗(yàn)結(jié)果和分析 22
第五章 性能評價(jià) 25
5.1 信噪比改進(jìn) 25
5.2 主觀評價(jià) 25
結(jié)論 27
參考文獻(xiàn) 29
參考文獻(xiàn)
[1] JAE.S.LIM,And V.Oppenbeim, Enhancement and Bandweidch Compression of Noise Speech[J].PROCEEDINGS OF THE IEEE,Vol.67
[2] Wang DL,Lim JS,The unimportance of phase in speech enhancement [J].IEEE Trans Acoust Speech Signal Processing,1982,30(4):679-681.
[3] Ephraim Y, Malah D. Speech enhancement using a minimum mean-square error short-time spectra amplitude estimator [J].IEEE Trans Acoust Speech Signal Processing,1984.32(6):1109-1121.
[4] S. F. Boll, “Suppression of acoustic noise in speech using spectral subtraction”, IEEE
Transaction on Acoustics, Speech and Signal Proc., Vol. 27, pp. 113–120,1979
[5] M. Berouti, R. Schwartz, J. Makhoul, “Enhancement of speech corrupted by acoustic noise”, Proc. IEEE ICASSP, pp.208-211, Washington, DC, April 1979
[6] P. Lockwood, J. Boudy, “Experiments with a nonlinear spectral subtraction(NSS),hidden
Markov models and projection for robust recognition in cars”, Speech Communication,
pp.215-228, November 1992
[7] B. Lim Sim, Y. C. Tong, “A parametric formulation of the generalized spectral
subtraction method”, IEEE Transaction on Speech and Audio Processing,pp.328-337,
June 1998
[8] N. Virag, “Single channel speech enhancement based on masking properties of human
auditory system”, IEEE Transactions on Speech and Audio Processing, pp.126-137, July
1999
[9] “JAE.S.LIM,And V.Oppenbeim, Enhancement and Bandweidch Compression of Noise Speech[J].PROCEEDINGS OF THE IEEE,Vol,67(12):1586-1604”
[10] 張波等,“低信噪比下的一種自適應(yīng)有聲/無聲判決算法”,信號處理,1996
[11] 沈亞強(qiáng),程仲文。一種基于自適應(yīng)的語音增強(qiáng)方法[J]。信號處理,1993,9(1):9-14
[12] 付強(qiáng),易克初,田斌,語音信號處理[M] 國防工業(yè)出版社
[13] 張金杰,曹志剛等。一種基于聽覺巖壁效應(yīng)的語音增強(qiáng)方法。清華大學(xué)學(xué)報(bào)(自然科學(xué)版),2001,41(7):1-4
[14] 廖春波。語音增強(qiáng)及其相關(guān)技術(shù)的研究。[大連理工大學(xué)學(xué)位論文]2003
[15] Tsoukalas D E, Mourjopoulos J N. Speech enhancement based on audible noise suppression.IEEETrans Speech and AudioProcessing,1997,5(6)497-514
[16] 劉海濱,“非平穩(wěn)環(huán)境下基于人耳聽覺掩蔽特性的語音增強(qiáng)”,信號處理,Vol.19,No.4,2003
[17] 王永紅等,“譜相減技術(shù)中“音樂噪聲”的抑制”,桂林電子工業(yè)學(xué)院學(xué)報(bào),
Vol.18.No.1,1998
摘 要
語音增強(qiáng)的目的主要是改進(jìn)語音質(zhì)量,在消除背景噪音的同時提高語音可懂度,但是這兩個目的往往不能兼得。目前有一些對非平穩(wěn)噪聲干擾下的語音信號進(jìn)行增強(qiáng)的方法,可以降低背景噪聲,但有時會引入刺耳的音樂噪聲,且不能提高語音的可懂度,甚至略有下降。其中譜減法作為一種單信道語音增強(qiáng)方法,以其簡單有效在使用化的語音增強(qiáng)應(yīng)用中深受歡迎。
本文研究采用閥值法法對非平穩(wěn)背景噪聲信號進(jìn)行估計(jì),計(jì)算出先驗(yàn)信噪比,得到還原的純凈語音信號。本文用MATLAB實(shí)現(xiàn)了整個算法的仿真,并與傳統(tǒng)譜減法結(jié)果相比較,仿真結(jié)果表明,該算法對非平穩(wěn)噪聲追蹤性較好,在抑制背景噪聲,減少音樂噪聲前提下,提高了語音的可懂度,其計(jì)算復(fù)雜度也可以接受。
關(guān)鍵詞: 譜減法 語音增強(qiáng) 噪聲估計(jì)
Abstract
The mainly purpose of speech enhancement is to improve the quality of speech, raises the intelligibility of speech while eliminating background noise, but this two purposes can not often be gotten simultaneously. Now there are many method of speech enhancement for the speech signal with near-stationary noise can reduce background noise, but can produce an annoying noise called music noise, and the intelligibility can not raise even drops. As a method of one channel speech enhancement, spectral subtraction is deeply welcome because it is easy and efficient.
In this paper, the SNR of the speech signal is estimated, then the enhanced speech signal is gotten. MATLAB is used to realize the purpose, and it is compared with traditional spectral subtraction. It shows that this algorithm is better for the speech signal with the near-stationary noise; it restrains the music noise and reduces the background noise, it also raises the intelligibility. The complex degree also can be accepted.
Keyword: Spectral subtraction Speech enhancement
Noise estimation
目錄
第一章 緒論 1
1.1 課題背景及問題提出 1
1.2 語音增強(qiáng)研究的主要內(nèi)容和發(fā)展方向 1
1.3 譜減法語音增強(qiáng)的研究現(xiàn)狀 3
第二章 語音感知和噪聲特性 5
2.1 語音特性 5
2.2噪聲特性 5
第三章 譜減法語音增強(qiáng)算法 7
3.1 語音增強(qiáng)算法概述 7
3.2一般譜減法 8
3.2.1 基本原理 8
3.2.3 基本步驟和仿真結(jié)果 11
3. 3 “音樂噪聲”的產(chǎn)生 13
3.4 改進(jìn)的譜減法 14
3.4.1改進(jìn)型譜減法的種類 14
3.4.2 幅度譜平滑 15
3.4.3 被減項(xiàng)加權(quán)值處理 16
3.4.4 功率譜修正處理 16
第四章 基于先驗(yàn)信噪比估計(jì)的改進(jìn)型譜減法 19
4.1 基本原理 19
4.2 寂靜幀和語音幀的判定 19
4.3 算法的實(shí)現(xiàn) 21
4.4 參數(shù)的估計(jì) 22
4.5 實(shí)驗(yàn)結(jié)果和分析 22
第五章 性能評價(jià) 25
5.1 信噪比改進(jìn) 25
5.2 主觀評價(jià) 25
結(jié)論 27
參考文獻(xiàn) 29
參考文獻(xiàn)
[1] JAE.S.LIM,And V.Oppenbeim, Enhancement and Bandweidch Compression of Noise Speech[J].PROCEEDINGS OF THE IEEE,Vol.67
[2] Wang DL,Lim JS,The unimportance of phase in speech enhancement [J].IEEE Trans Acoust Speech Signal Processing,1982,30(4):679-681.
[3] Ephraim Y, Malah D. Speech enhancement using a minimum mean-square error short-time spectra amplitude estimator [J].IEEE Trans Acoust Speech Signal Processing,1984.32(6):1109-1121.
[4] S. F. Boll, “Suppression of acoustic noise in speech using spectral subtraction”, IEEE
Transaction on Acoustics, Speech and Signal Proc., Vol. 27, pp. 113–120,1979
[5] M. Berouti, R. Schwartz, J. Makhoul, “Enhancement of speech corrupted by acoustic noise”, Proc. IEEE ICASSP, pp.208-211, Washington, DC, April 1979
[6] P. Lockwood, J. Boudy, “Experiments with a nonlinear spectral subtraction(NSS),hidden
Markov models and projection for robust recognition in cars”, Speech Communication,
pp.215-228, November 1992
[7] B. Lim Sim, Y. C. Tong, “A parametric formulation of the generalized spectral
subtraction method”, IEEE Transaction on Speech and Audio Processing,pp.328-337,
June 1998
[8] N. Virag, “Single channel speech enhancement based on masking properties of human
auditory system”, IEEE Transactions on Speech and Audio Processing, pp.126-137, July
1999
[9] “JAE.S.LIM,And V.Oppenbeim, Enhancement and Bandweidch Compression of Noise Speech[J].PROCEEDINGS OF THE IEEE,Vol,67(12):1586-1604”
[10] 張波等,“低信噪比下的一種自適應(yīng)有聲/無聲判決算法”,信號處理,1996
[11] 沈亞強(qiáng),程仲文。一種基于自適應(yīng)的語音增強(qiáng)方法[J]。信號處理,1993,9(1):9-14
[12] 付強(qiáng),易克初,田斌,語音信號處理[M] 國防工業(yè)出版社
[13] 張金杰,曹志剛等。一種基于聽覺巖壁效應(yīng)的語音增強(qiáng)方法。清華大學(xué)學(xué)報(bào)(自然科學(xué)版),2001,41(7):1-4
[14] 廖春波。語音增強(qiáng)及其相關(guān)技術(shù)的研究。[大連理工大學(xué)學(xué)位論文]2003
[15] Tsoukalas D E, Mourjopoulos J N. Speech enhancement based on audible noise suppression.IEEETrans Speech and AudioProcessing,1997,5(6)497-514
[16] 劉海濱,“非平穩(wěn)環(huán)境下基于人耳聽覺掩蔽特性的語音增強(qiáng)”,信號處理,Vol.19,No.4,2003
[17] 王永紅等,“譜相減技術(shù)中“音樂噪聲”的抑制”,桂林電子工業(yè)學(xué)院學(xué)報(bào),
Vol.18.No.1,1998