Statistical-model-based speech enhancement systems

Ephraim, Yariv

doi:10.1109/5.168664

Cited by 320 publications

(122 citation statements)

References 144 publications

Supporting

Mentioning

116

Contrasting

Unclassified

Order By: Relevance

“…Several approaches were studied for speech enhancement in additive noise (Boll, 1979;Gibson et al, 1991;Cheng and O'Shaughnessy, 1991;Ephraim, 1992;Ephraim and Van Trees, 1995;Le Bouquin, 1996;Lee and Shirai, 1996). Many of these studies have focussed on enhancement based www.elsevier.nl/locate/specom Speech Communication 28 (1999) 25±42 on attempts to suppress noise (Boll, 1979;Ephraim and Van Trees, 1995;Le Bouquin, 1996).…”

Section: Introductionmentioning

confidence: 99%

Speech enhancement using linear prediction residual

Yegnanarayana

Avendaño

Heřmanský

et al. 1999

Speech Communication

View full text Add to dashboard Cite

In this paper we propose a method for enhancement of speech in the presence of additive noise. The objective is to selectively enhance the high signal-to-noise ratio (SNR) regions in the noisy speech in the temporal and spectral domains, without causing signi®cant distortion in the resulting enhanced speech. This is proposed to be done at three di erent levels. (a) At the gross level, by identifying the regions of speech and noise in the temporal domain. (b) At the ®ner level, by identifying the regions of high and low SNR portions in the noisy speech. (c) At the short-time spectrum level, by enhancing the spectral peaks over spectral valleys. The basis for the proposed approach is to analyze linear prediction (LP) residual signal in short (1±2 ms) segments to determine whether a segment belongs to a noise region or speech region. In the speech regions the inverse spectral¯atness factor is signi®cantly higher than in the noisy regions. The LP residual signal enables us to deal with short segments of data due to uncorrelatedness of the samples. Processing of noisy speech for enhancement involves mostly weighting the LP residual signal samples. The weighted residual signal samples are used to excite the time-varying all-pole ®lter to produce enhanced speech. As the additive noise level in the speech signal is increased, the quality of the resulting enhanced speech decreases progressively due to loss of speech information in the low SNR, high noise regions. Thus the degradation in performance of enhancement is graceful as the overall SNR of the noisy speech is decreased. Ó

show abstract

Section: Introductionmentioning

confidence: 99%

Speech enhancement using linear prediction residual

Yegnanarayana

Avendaño

Heřmanský

et al. 1999

Speech Communication

View full text Add to dashboard Cite

show abstract

“…Parameter values of these models are estimated from samples of speech in the testing environments, and either the features of the incoming speech or the internally-stored representations of speech in the system are modified. Typical structural models for adaptation to acoustical variability assume that speech is corrupted either by additive noise with an unknown power spectrum (Porter & Boll, 1984;Ephraim, 1992;Erell & Weintraub, 1990;Gales & Young, 1992;Lockwood, Boudy, et al, 1992;Bellegarda, de Souza, et al, 1992), or by a combination of additive noise and linear filtering (Acero & Stern, 1990). Much of the early work in robust recognition involved a re-implementation of techniques developed to remove additive noise for the purpose of speech enhancement, as reviewed in section 10.3.…”

Section: Optimal Parameter Estimationmentioning

confidence: 99%

Survey of the State of the Art in Human Language Technology

Cohen¹,

Varile²,

Zampolli³

et al. 2000

Language

133

View full text Add to dashboard Cite

“…Time domain techniques were based on conventional filtering approach such as linear predictive coding based filtering, Hidden Markov Model and Kalman filtering. Hidden Markov Model is used firstly for speech enhancement [9][10][11][12][13][14][15][16][17]. A marked research progress have been made in many transformation methods such as Discrete Cosine Transform (DCT), Fast Fourier Transform (FFT), Karhunen-Loeve Transform (KLT) Discrete Wavelet Transform (DWT), Ridgelet and Curvelet transform which are extensively used in data compression, detection and classification.…”

Section: Introductionmentioning

confidence: 99%

2-D Speech Enhancement based on Curvelet Transform using Different Window Functions

Verma¹,

Verma²,

Kumar³

2013

IJCA

View full text Add to dashboard Cite

In this paper, an improved method based on Curvelet Transform using different window functions is presented for the speech enhancement. The window function is used for preprocessing of speech signals. In this method, instead of using two-dimensional (2-D) discrete Fourier Transform, Curvelet transform is employed with spectral magnitude subtraction method. General termsSpectral Substraction Method, Curvlet transform.

show abstract

Statistical-model-based speech enhancement systems

Cited by 320 publications

References 144 publications

Speech enhancement using linear prediction residual

Speech enhancement using linear prediction residual

Survey of the State of the Art in Human Language Technology

2-D Speech Enhancement based on Curvelet Transform using Different Window Functions

Contact Info

Product

Resources

About