基于噪聲被掩蔽概率的優(yōu)化語音增強方法
-
1.
北京大學信息科學技術(shù)學院 北京 100871
-
2.
西安交通大學生命科學技術(shù)學院 西安 710049
-
3.
西安通信學院計算中心 西安 710106
Optimizing Speech Enhancement Based on Noise Masked Probability
-
摘要: 利用聽覺系統(tǒng)的掩蔽特性,提出了一種優(yōu)化的語音增強方法。研究表明,噪聲被語音掩蔽的概率是噪聲強度和聽覺掩蔽閾值的函數(shù)??紤]到噪聲在帶噪語音中的出現(xiàn)具有不確定性,各語音譜分量的最終估計由對帶噪語音的譜分量和用傳統(tǒng)的增強方法估計的譜分量的加權(quán)求得,加權(quán)因子由噪聲被掩蔽概率確定。語音增強性能的評估結(jié)果表明,這種優(yōu)化的語音增強方法在減少語音失真與加強噪聲抑制之間取得了良好的折衷,減少了語音的聽覺失真, 有效地抑制了音樂噪聲,提高了增強語音的清晰度。
-
關(guān)鍵詞:
- 語音增強; 聽覺掩蔽效應; 語音清晰度; 音樂噪聲
Abstract: An optimal approach for enhancing a speech signal degraded by uncorrelated stationary additive noise, which exploits auditory perception properties, is proposed. The speech spectra estimate is performed in two cases: noisy speech spectra for noise masked and classical estimate for noise unmasked. Taking account into the uncertainty of the noise presence, the enhanced speech signal spectra are obtained by a weighted sum of these two estimates, where the weights are given by the noise masked probability. The performance of the proposed speech enhancement approach has been evaluated with speech distortion and informal listening tests. Comparing with Aziranis method and classical estimator, results show that a better compromise between reducing speech distortion and reinforcing noise suppression has been made, speech distortion has been decreased apparently, musical noise has been suppressed and speech articulation has been improved. -
Ephraim Y, Malah D. Speech enhancement using a minimum mean square error short-time spectral amplitude estimator. IEEE Trans. onASSP, 1984, 32(6): 1109- 1121.[2]Cappe O. Elimination of the musical noise phenomenon with the Ephraim and Malah noise suppressor[J].IEEE Trans. on Speech and Audio Processing.1994, 2(3):345-[3]McAulay R J, Malpass M L. Speech enhancement using a soft decision noise suppression filter[J].IEEE Trans. on ASSP.1980,28(2):137-[4]曹志剛,鄭文濤.基于短時譜最小均方誤差估計的語音增強和剩余噪聲衰減.電子學報,1993,21(4):7-12.[5]陸生禮,時龍興,余崇智,等.聽覺模擬的語音增強方法.聲學學報,1996,21(6):879-883.[6]Virag N. Single channel speech enhancement based on masking properties of the human auditory system[J].IEEE Trans. on Speech and Audio Processing.1999, 7(2):126-[7]Tsoukalas D E, M Paraskevas, Mourjopoulos J N. Speech enhancement using psychoacoustic criteria. ICASSP, Salt Lake City, 1993, Ⅱ: 359 - 362.[8]Azirani A, Jeannes R L B, Faucon G. Optimizing speech enhancement by exploiting masking properties of the human ear.ICASSP, Detroit, 1995, I: 800 - 803.[9]沈永歡,梁在中,許履瑚,等.實用數(shù)學手冊.北京:科學出版社,1997:477-479.[10]Bu Fanliang, Hou Zhen, Wen Yuan, et al.. An estimation of noise parameters in noisy speech signals. NCMMSC6, Shenzhen, 2001:71 - 74.[11]Johnston J D. Transform coding of audio signal using perceptual noise criteria[J].IEEE J. on Select. Areas Commun.1988, 6(2):314- -
計量
- 文章訪問數(shù): 2648
- HTML全文瀏覽量: 78
- PDF下載量: 814
- 被引次數(shù): 0