Paper Submission
| 4/2011 - 1 | View TOC | « Previous Article | Next Article » |
Line Spectral Frequency-based Noise Suppression for Speech-Centric Interface of Smart DevicesJANG, G. J.   , PARK, J. S. , KIM, J. H. , SEO, Y. H. |
| Click to see author's profile on SCOPUS, IEEE Xplore, Web of Science |
Download PDF  (708 KB) | Citation | Downloads: 1,565 | Views: 2,734 |
Author keywords
noise measurement, noise reduction, speech enhancement, speech recognition, linear predictive coding
References keywords
speech(15), spectral(6), processing(5), signal(4), recognition(4)
Blue keywords are present in both the references section and the paper title.
About this article
Date of Publication: 2011-11-30
Volume 11, Issue 4, Year 2011, On page(s): 3 - 8
ISSN: 1582-7445, e-ISSN: 1844-7600
Digital Object Identifier: 10.4316/AECE.2011.04001
Web of Science Accession Number: 000297764500001
SCOPUS ID: 84863083144
Abstract
This paper proposes a noise suppression technique for speech-centric interface of various smart devices. The proposed method estimates noise spectral magnitudes from line spectral frequencies (LSFs), using the observation that adjacent LSFs correspond to peak frequencies of spectrum, whereas isolated LSFs are close to flattened valley frequencies retaining noise components. Over a course of segmented time frames, the logarithms of spectral magnitudes at respective LSFs are computed, and their distribution is then modeled by the Rayleigh probability density function. The standard deviation from the Rayleigh function approximates the noise spectral magnitude. The model is updated at every frame in an online manner so that it can deal with real-time inputs. Once the noise spectral magnitude is estimated, a time-domain Wiener filter is derived for the suppression of the estimated noise spectral magnitude, and this is then applied to the input noisy speech signals. Our proposed approach operates well on most smart devices owing to its low computational complexity and real-time implementation. Speech recognition experiments, conducted to evaluate the proposed technique, show that our method exhibits superior performance, with less distortion of original speech, when compared to conventional noise suppression techniques. |
| References | | | Cited By «-- Click to see who has cited this paper |
| [1] M. Schuricht, Z. Davis, M. Hu, S. Prasad, P. Melliar-Smith, and L. Moser, "Managing multiple speech-enabled applications in a mobile handheld device," International Journal of Pervasive Computing and Communications, vol. 5, no. 3, pp. 332-359, Sep. 2009. [CrossRef] [SCOPUS Times Cited 3] [2] L. Deng, A. Acero, Y. Wang, K. Wang, H. Hon, et al., "A speech-centric perspective for human-computer interface," IEEE Workshop on Multimedia Signal Processing, pp. 263-267, Dec. 2002. [CrossRef] [SCOPUS Times Cited 3] [3] K. Kim and M. Kim, "Robust speaker recognition against background noise in an enhanced multi-condition domain," IEEE Transactions on Consumer Electronics, vol. 56, no. 3, pp. 1684-1688, Aug. 2010. [CrossRef] [Web of Science Times Cited 6] [SCOPUS Times Cited 8] [4] S. Boll, "Suppression of acoustic noise in speech using spectral subtraction," IEEE Transactions on Acoustics, Speech and Signal Processing, vol. 27, no. 2, pp. 113-120, Apr. 1979. [CrossRef] [Web of Science Times Cited 1645] [SCOPUS Times Cited 2370] [5] K. Wu and P. Chen, "Efficient speech enhancement using spectral subtraction for car hands-free applications," Proc. of International Conference on Consumer Electronics, pp. 220-221, Jun. 2001. [CrossRef] [Web of Science Record] [6] V. Stahl, A. Fischer, and R. Bippus, "Quantile based noise estimation for spectral subtraction and wiener filtering," Proc. of ICASSP, vol. 3, pp. 1875-1878, Jun. 2000. [CrossRef] [7] A. Kindoz and A. Kondoz, Digital speech; coding for low bit rate communication systems, John Wiley & Sons, Inc., New York, NY, USA, Jan. 1994. [8] P. Kabal and R. Ramachandran, "The computation of line spectral frequencies using chebyshev polynomials," IEEE Transactions on Acoustics, Speech, Signal Processing, vol. 34, no. 6, pp. 1419-1426, Dec. 1986. [CrossRef] [Web of Science Times Cited 81] [SCOPUS Times Cited 98] [9] M. Lee, H. Kim, S. Choi, and H. Lee, "On the use of LSF intermodel interlacing property for spectral quantization," Proc. of IEEE Workshop on Speech Coding, pp. 43-45, Jun. 1999. [CrossRef] [10] M. Lee, H. Kim, and H. Lee, "A new distortion measure for spectral quantization based on the LSF intermodel interlacing property," Speech Communication, vol. 35, no. 3-4, pp. 191-202, Oct. 2001. [CrossRef] [Web of Science Times Cited 6] [SCOPUS Times Cited 7] [11] T. Backstrom and C. Magi, "Properties of line spectrum pair polynomials - a review," Signal Processing, vol. 86, pp. 3286-3298, Nov. 2006. [CrossRef] [Web of Science Times Cited 18] [SCOPUS Times Cited 24] [12] Telecommunications Industry Association (TIA), "Enhanced variable rate codec, speech service option 3 for wideband spread spectrum digital systems," Technical Report, TIA/EIA/IS-127-2, Dec. 1999. [13] European Telecommunications Standards Institute, "Speech processing, transmission and quality aspects (STQ); distributed speech recognition; advanced front-end feature extraction algorithm; compression algorithm," Technical Report, ES 202 050 v1.1.5, Jan. 2007. [14] M. Cooke, J. Hershey, and S. Rennie, "Monaural speech separation and recognition challenge," Computer Speech & Language, vol. 24, no. 1, pp. 1-15, 2010. [CrossRef] [Web of Science Times Cited 72] [SCOPUS Times Cited 90] [15] S. Young, G. Evermann, M. Gales, T. Hain, D. Kershaw, X. Liu, et al., Hidden Markov model toolkit (HTK), ver. 3.4, Dec. 2006. [Online] Available: Temporary on-line reference link removed - see the PDF document [16] D. Pearce and H. Hirsch, "The AURORA experimental framework for the performance evaluations of speech recognition systems under noisy condition," Proc. of ICSLP, Oct. 2000. [Online] Available: Temporary on-line reference link removed - see the PDF document Web of Science® Citations for all references: 1,828 TCR SCOPUS® Citations for all references: 2,603 TCR Web of Science® Average Citations per reference: 114 ACR SCOPUS® Average Citations per reference: 163 ACR TCR = Total Citations for References / ACR = Average Citations per Reference We introduced in 2010 - for the first time in scientific publishing, the term "References Weight", as a quantitative indication of the quality ... Read more Citations for references background updated on 2017-03-05 12:39 in 75 seconds. Note1: Web of Science® is a registered trademark of Thomson Reuters. Note2: SCOPUS® is a registered trademark of Elsevier B.V. Disclaimer: All queries to the respective databases were made by using the DOI record of every reference (where available). Due to technical problems beyond our control, the information is not always accurate. Please use the CrossRef link to visit the respective publisher site. |
Copyright ©2001-2017
Faculty of Electrical Engineering and Computer Science
Stefan cel Mare University of Suceava, Romania
All rights reserved: Advances in Electrical and Computer Engineering is a registered trademark of the Stefan cel Mare University of Suceava. No part of this publication may be reproduced, stored in a retrieval system, photocopied, recorded or archived, without the written permission from the Editor. When authors submit their papers for publication, they agree that the copyright for their article be transferred to the Faculty of Electrical Engineering and Computer Science, Stefan cel Mare University of Suceava, Romania, if and only if the articles are accepted for publication. The copyright covers the exclusive rights to reproduce and distribute the article, including reprints and translations.
Permission for other use: The copyright owner's consent does not extend to copying for general distribution, for promotion, for creating new works, or for resale. Specific written permission must be obtained from the Editor for such copying. Direct linking to files hosted on this website is strictly prohibited.
Disclaimer: Whilst every effort is made by the publishers and editorial board to see that no inaccurate or misleading data, opinions or statements appear in this journal, they wish to make it clear that all information and opinions formulated in the articles, as well as linguistic accuracy, are the sole responsibility of the author.
Faculty of Electrical Engineering and Computer Science
Stefan cel Mare University of Suceava, Romania
All rights reserved: Advances in Electrical and Computer Engineering is a registered trademark of the Stefan cel Mare University of Suceava. No part of this publication may be reproduced, stored in a retrieval system, photocopied, recorded or archived, without the written permission from the Editor. When authors submit their papers for publication, they agree that the copyright for their article be transferred to the Faculty of Electrical Engineering and Computer Science, Stefan cel Mare University of Suceava, Romania, if and only if the articles are accepted for publication. The copyright covers the exclusive rights to reproduce and distribute the article, including reprints and translations.
Permission for other use: The copyright owner's consent does not extend to copying for general distribution, for promotion, for creating new works, or for resale. Specific written permission must be obtained from the Editor for such copying. Direct linking to files hosted on this website is strictly prohibited.
Disclaimer: Whilst every effort is made by the publishers and editorial board to see that no inaccurate or misleading data, opinions or statements appear in this journal, they wish to make it clear that all information and opinions formulated in the articles, as well as linguistic accuracy, are the sole responsibility of the author.