| 2/2022 - 9 |
A Novel Approach to Speech Enhancement Based on Deep Neural NetworksSALEHI, M.   , MIRZAKUCHAKI, S. |
View the paper record and citations in  |
| Click to see author's profile in SCOPUS, IEEE Xplore, Web of Science |
Download PDF  (3,059 KB) | Citation | Downloads: 657 | Views: 1,129 |
Author keywords
long short-term memory, machine learning, mean square error methods, recurrent neural networks, speech enhancement
References keywords
speech(31), processing(15), access(14), enhancement(13), learning(11), signal(9), noise(9), estimation(8), deep(8), spectral(7)
Blue keywords are present in both the references section and the paper title.
About this article
Date of Publication: 2022-05-31
Volume 22, Issue 2, Year 2022, On page(s): 71 - 78
ISSN: 1582-7445, e-ISSN: 1844-7600
Digital Object Identifier: 10.4316/AECE.2022.02009
Web of Science Accession Number: 000810486800009
SCOPUS ID: 85131727505
Abstract
Minimum mean-square error (MMSE) approaches have been shown to achieve state-of-the-art performance on the task of speech enhancement. However, MMSE approaches lack the ability to accurately estimate non-stationary noise sources. In this paper, a long short-term memory fully convolutional network (LSTM-FCN) is utilized to accurately estimate a priori signal-to-noise ratio (SNR) since the speech enhancement performance of an MMSE approach improves with the accuracy of the used a priori SNR estimator. The proposed MMSE approach makes no assumptions about the characteristics of the noise or the speech. MMSE approaches that utilize the LSTM-FCN estimator are evaluated using the mean opinion score of the perceptual evaluation of speech quality (PESQ) and the short-time objective intelligibility (STOI) measures of speech. The experimental investigation shows that the speech enhancement performance of an MMSE approach that utilizes LSTM-FCN estimator significantly increases. |
| References | | | Cited By «-- Click to see who has cited this paper |
| [1] S. K. Roy, A. Nicolson, K. K. Paliwal, "DeepLPC: A deep learning approach to augmented Kalman filter-based single-channel speech enhancement," IEEE Access, vol. 9, no. 4, pp. 64524-64538, 2021. [CrossRef] [Web of Science Times Cited 10] [2] Z. Q. Wang, P. Wang, D. Wang, "Complex spectral mapping for single-and multi-channel speech enhancement and robust ASR," IEEE/ACM Trans. Audio, Speech, and Language Processing, vol. 28, no. 5, pp. 1778 -1787, 2020. [CrossRef] [Web of Science Times Cited 99] [3] S. Othman, A. Mohamed, A. Abouali, Z. Nossair, "Lossy compression using adaptive polynomial image encoding," Advances in Electrical and Computer Engineering, vol.21, no.1, pp.91-98, 2021. [CrossRef] [Full Text] [Web of Science Times Cited 4] [4] T. G. Yadava, H. S. Jayanna, "Speech enhancement by combining spectral subtraction and minimum mean square error-spectrum power estimator based on zero crossing," International Journal of Speech Technology, vol. 22, no. 3, pp. 639-648, 2019. [CrossRef] [Web of Science Times Cited 14] [5] Y. Zhang, Y. Zhao, "Real and imaginary modulation spectral subtraction for speech enhancement," Speech Communication, vol. 55, no. 4, pp. 509-522, 2013. [CrossRef] [Web of Science Times Cited 30] [6] Y. Ephraim, D. Malah, "Speech enhancement using a minimum-mean square error short-time spectral amplitude estimator," IEEE Trans. on Acoustics, Speech, and Signal processing, vol. 32, no. 6, pp. 1109-1121, 1984. [CrossRef] [Web of Science Times Cited 2516] [7] P. C. Loizou, Speech enhancement: Theory and practice. CRC press, 2007 [8] B. M. Mahmmod, A. R. Ramli, S. H. Abdulhussian, S. A. R. Al-Haddad, W. A. Jass, "Low-distortion MMSE speech enhancement estimator based on Laplacian Prior," IEEE Access, vol. 5, no. 4, pp. 9866-9881, 2017. [CrossRef] [Web of Science Times Cited 30] [9] A. Brown, S. Garg, J. Montgomery, "Automatic and efficient denoising of bioacoustics recordings using MMSE STSA," IEEE Access, vol. 6, no. 12, pp. 5010-5022, 2017. [CrossRef] [Web of Science Times Cited 14] [10] Q. Zhang, M. Wang, Y. Lu, M. Idrees, L. Zhang, "Fast nonstationary noise tracking based on log-spectral power MMSE estimator and temporal recursive averaging," IEEE Access, vol. 7, no. 6, pp. 80985-80999, 2019. [CrossRef] [Web of Science Times Cited 6] [11] C. Plapous, C. Marro, L. Mauuary, P. Scalart, "A two-step noise reduction technique," in IEEE International Conf. on Acoustics, Speech, and Signal Processing, Montreal, 2004, pp. 289-292. [CrossRef] [12] C. Plapous, C. Marro, P. Scalart, "Improved signal-to-noise ratio estimation for speech enhancement," IEEE Trans. on Audio, Speech, and Language Processing, vol. 14, no. 6, pp. 2098-2108, 2006. [CrossRef] [Web of Science Times Cited 211] [13] Y. G. Thimmaraja, B. Nagaraja, H. Jayanna, "Speech enhancement and encoding by combining SS-VAD and LPC," International Journal of Speech Technology, vol. 24, no. 1, pp. 165-172, 2021. [CrossRef] [Web of Science Times Cited 9] [14] F. Bellili, R. Meftehi, S. Affes, A. Stephenne, "Maximum likelihood SNR estimation of linearly-modulated signals over time-varying flat-fading SIMO channels," IEEE Trans. on Signal Processing, vol. 63, no. 2, pp. 441-456, 2014. [CrossRef] [Web of Science Times Cited 30] [15] C. Breithaupt, T. Gerkmann, R. Martin, "A novel a priori SNR estimation approach based on selective cepstro-temporal smoothing," in IEEE International Conf. on Acoustics, Speech and Signal Processing, Las Vegas, 2008, pp. 4897-4900. [CrossRef] [Web of Science Times Cited 92] [16] V. Timcenko, S. Gajin, "Machine learning enhanced entropy-based network anomaly detection," Advances in Electrical and Computer Engineering, vol.21, no.4, pp.51-60, 2021. [CrossRef] [Full Text] [17] A. Albu, R. E Precup, T. A Teban, "Results and challenges of artificial neural networks used for decision-making in medical applications," FACTA Universitatis Series: Mechanical Engineering, vol. 17, no. 3, pp. 285-308, 2019. [CrossRef] [Web of Science Times Cited 96] [18] T. Zhang, F. Xu, T. Wu, "A software tool for spiking neural P systems," Romanian Journal of Information Science and Technology, vol. 23, no. 1, pp. 84-92, 2020 [19] E. L. Hedrea, R. E. Precup, R. C. Roman, E. M. Petriu, "Tensor product-based model transformation approach to tower crane systems modeling," Asian Journal of Control, vol. 23, no. 3, pp. 1313-1323, 2021. [CrossRef] [Web of Science Times Cited 71] [20] J. B. Awotunde, R. O. Ogundokun, F. E. Ayo, O. E. Matiluko, "Speech segregation in background noise based on deep learning," IEEE Access, vol. 8, no. 9, pp. 169568-169575, 2020. [CrossRef] [Web of Science Times Cited 6] [21] J. Kim, M. Hahn, "Speech enhancement using a two-stage network for an efficient boosting strategy," IEEE Signal Processing Letters, vol. 26, no. 5, pp. 770-774, 2019. [CrossRef] [Web of Science Times Cited 11] [22] X. L. Zhang, D. Wang, "Boosting contextual information for deep neural network based voice activity detection," IEEE/ACM Trans. on Audio, Speech, and Language Processing, vol. 24, no. 2, pp. 252-264, 2015. [CrossRef] [Web of Science Times Cited 104] [23] Q. Wang, J. Du, L. R. Dai, C. H. Lee, "A multiobjective learning and ensembling approach to high-performance speech enhancement with compact neural network architectures," IEEE/ACM Transactions on Audio, Speech, and Language Processing, vol. 26, no. 7, pp. 1185-1197, 2018. [CrossRef] [Web of Science Times Cited 33] [24] I. Ahmed, S. Alam, J. Hossain, G. Kaddoum, "Deep learning for MMSE estimation of a Gaussian source in the presence of Bursty impulsive noise," IEEE Communications Letters, vol. 25, no. 4, pp. 1211-1215, 2020. [CrossRef] [Web of Science Times Cited 5] [25] A. Nicolson, K. K. Paliwal, "Deep learning for minimum mean-square error approaches to speech enhancement," Speech Communication, vol. 111, no. 8, pp. 44-55, 2019. [CrossRef] [Web of Science Times Cited 78] [26] Q. Zhang, A. Nicolson, M. Wang, K. K. Paliwal, C. Wang, "DeepMMSE: A deep learning approach to MMSE-based noise power spectral density estimation," IEEE/ACM Trans. on Audio, Speech, and Language Processing, vol. 28, no. 4, pp. 1404-1415, 2020. [CrossRef] [Web of Science Times Cited 70] [27] F. Karim, S. Majumdar, H. Darabi, S. Chen, "LSTM fully convolutional networks for time series classification," IEEE Access, vol. 6, no. 12, pp. 1662-1669, 2017. [CrossRef] [Web of Science Times Cited 682] [28] F. Karim, S. Majumdar, H. Darabi, "Insights into LSTM fully convolutional networks for time series classification," IEEE Access, vol.7, no. 5, pp. 67718-67725, 2019. [CrossRef] [Web of Science Times Cited 107] [29] R. A. Yu, "A low-complexity noise estimation algorithm based on smoothing of noise power estimation and estimation bias correction," in IEEE International Conf. on Acoustics, Speech and Signal Processing, Taipei, 2009, pp. 4421-4424. [CrossRef] [30] R. C. Hendriks, R. Heusdens, J. Jensen, "MMSE based noise PSD tracking with low complexity," in IEEE International Conf. on Acoustics, Speech and Signal Processing, Dallas, 2010, pp. 4266-4269. [CrossRef] [Web of Science Times Cited 174] [31] Y. Ephraim, D. Malah, "Speech enhancement using a minimum mean-square error log-spectral amplitude estimator," IEEE Trans. on Acoustics, Speech, and Signal Processing, vol. 33, no. 2, pp. 443-445, 1985. [CrossRef] [Web of Science Times Cited 1397] [32] I. Goodfellow I, Y. Bengio, A. Courville, Deep learning. MIT press, pp. 257-267, 2016 [33] T. Y. Hsiao, Y. C. Chang, H. H. Chou, C. T. Lin, "Filter-based deep-compression with global average pooling for convolutional networks," Journal of Systems Architecture, vol. 95, no. 5, pp. 9-18, 2019. [CrossRef] [Web of Science Times Cited 55] [34] C. LeaEmail, R. Vidal, A. Reiter, G. D. Hager, "Temporal Convolutional Networks: A unified approach to action segmentation," in European Conf. on Computer Vision, Amsterdam, 2016, pp. 47-54. [CrossRef] [Web of Science Times Cited 392] [35] P. Dhruv, S. Naskar, "Image classification using convolutional neural network (CNN) and recurrent neural network (RNN): A review," Machine Learning and Information Processing, vol. 1101, no. 3, pp. 367-381, 2020. [CrossRef] [36] X. Wu, X. Shen, J. Zhang, Y. Zhang, "A wind energy prediction scheme combining cauchy variation and reverse learning strategy," Advances in Electrical and Computer Engineering, vol.21, no.4, pp.3-10, 2021, [CrossRef] [Full Text] [37] A. Barakat, P. Bianchi, "Convergence and dynamical behavior of the ADAM algorithm for non-convex stochastic optimization," SIAM Journal on Optimization, vol. 31, no. 1, pp. 244-274, 2021. [CrossRef] [Web of Science Times Cited 31] [38] P. Netrapalli, "Stochastic gradient descent and its variants in machine learning," Journal of the Indian Institute of Science, vol. 99, no. 2, pp. 201-213, 2019. [CrossRef] [Web of Science Times Cited 40] [39] R. V. K. Reddy, B. S. Rao, K. P. Raju, "Handwritten Hindi digits recognition using convolutional neural network with RMSprop optimization," in Second International Conference on Intelligent Computing and Control Systems (ICICCS), Madurai, 2018, pp. 45-51. [CrossRef] [40] C. H. Taal, R. C. Hendriks, R. Heusdens, J. Jensen, "An algorithm for intelligibility prediction of time-frequency weighted noisy speech," IEEE Trans. on Audio, Speech, and Language Processing, vol. 17, no. 7, pp. 2125-2136, 2011. [CrossRef] [Web of Science Times Cited 1381] Web of Science® Citations for all references: 7,798 TCR SCOPUS® Citations for all references: 0 Web of Science® Average Citations per reference: 190 ACR SCOPUS® Average Citations per reference: 0 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 updated on 2024-04-23 01:35 in 209 seconds. Note1: Web of Science® is a registered trademark of Clarivate Analytics. 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 ©2001-2024
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.
