Abstract
The performance of many noise-robust automatic speech recognition (ASR) methods, such as vector Taylor series (VTS) feature compensation, heavily depends on an estimation of the noise that contaminates speech. Therefore, providing accurate noise estimates for this kind of methods is crucial as well as a challenge. In this paper we investigate the use of deep neural networks (DNNs) to perform noise estimation in dual-microphone smartphones. Thanks to the powerful regression capabilities of DNNs, accurate noise estimates can be obtained by just using simple features as well as exploiting the power level difference (PLD) between the two microphones of the smartphone when employed in close-talk conditions. This is confirmed by our word recognition results on the AURORA2-2C (AURORA2 - 2 Channels - Conversational Position) database by largely outperforming single- and dual-channel noise estimation algorithms from the state-of-the-art when used together with a VTS feature compensation method.
I. López-Espejo et al. This work has been supported by the Spanish MINECO TEC2013-46690-P project.
This is a preview of subscription content, log in via an institution to check access.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Similar content being viewed by others
References
Li, J., Deng, L., Gong, Y., Haeb-Umbach, R.: An overview of noise-robust automatic speech recognition. IEEE Trans. Audio Speech Lang. Process. 22(4), 745–777 (2014)
Moreno, P.J., et al.: A vector Taylor series approach for environment-independent speech recognition. In: ICASSP, Atlanta, USA, pp. 733–736 (1996)
Wu, J., Droppo, J., Deng, L., Acero, A.: A noise-robust ASR front-end using Wiener filter constructed from MMSE estimation of clean speech and noise. In: ASRU, Virgin Islands, pp. 321–326 (2003)
Rangachari, S., Loizou, P.C.: A noise-estimation algorithm for highly non-stationary environments. Speech Commun. 48, 220–231 (2006)
Cohen, I.: Noise spectrum estimation in adverse environments: IMCRA. IEEE Trans. Speech Audio Process. 11, 466–475 (2003)
Martin, R.: Noise power spectral density estimation based on optimal smoothing and minimum statistics. IEEE Trans. Speech Audio Process. 9, 504–512 (2001)
Hendriks, R.C., Heusdens, R., Jensen, J.: MMSE based noise PSD tracking with low complexity. In: ICASSP, Dallas, USA (2010)
Jeub, M., et al.: Noise reduction for dual-microphone mobile phones exploiting power level differences. In: ICASSP, Kyoto, Japan, pp. 1693–1696 (2012)
Zhang, J., et al.: A fast two-microphone noise reduction algorithm based on power level ratio for mobile phone. In: ISCSLP, Hong-Kong, pp. 206–209 (2012)
López-Espejo, I., et al.: Feature enhancement for robust speech recognition on smartphones with dual-microphone. In: EUSIPCO, Lisbon, Portugal (2014)
López-Espejo, I., González, J.A., Gómez, Á.M., Peinado, A.M.: A deep neural network approach for missing-data mask estimation on dual-microphone smartphones: application to noise-robust speech recognition. In: Navarro Mesa, J.L., Ortega, A., Teixeira, A., Hernández Pérez, E., Quintana Morales, P., Ravelo García, A., Guerra Moreno, I., Toledano, D.T. (eds.) IberSPEECH 2014. LNCS (LNAI), vol. 8854, pp. 119–128. Springer, Heidelberg (2014). doi:10.1007/978-3-319-13623-3_13
Wang, Y., Wang, D.L.: Towards scaling up classication-based speech separation. IEEE Trans. Audio Speech Lang. Process. 21(7), 1381–1390 (2013)
Vincent, E.: Is audio signal processing still useful in the era of machine learning? In: WASPAA, New York, USA (2015)
Hinton, G.E., Osindero, S., Teh, Y.W.: A fast learning algorithm for deep belief nets. Neural Comput. 18, 1527–1554 (2006)
Hinton, G.E., Salakhutdinov, R.: Reducing the dimensionality of data with neural networks. Science 313(5786), 504–507 (2006)
Seltzer, M.L., et al.: An investigation of deep neural networks for noise robust speech recognition. In: ICASSP, Vancouver, Canada, pp. 7398–7402 (2013)
Xu, Y., Du, J., Dai, L.R.: A regression approach to speech enhancement based on deep neural networks. IEEE Trans. Audio Speech Lang. Process. 23(1), 7–19 (2015)
Segura, J.C., et al.: Model-based compensation of the additive noise for continuous speech recognition. Experiments using the AURORA II database and tasks. In: EUROSPEECH, Aalborg, Denmark (2001)
Pearce, D., Hirsch, H.G.: The Aurora experimental framework for the performance evaluation of speech recognition systems under noisy conditions. In: ICSLP, Beijing, China (2000)
ETSI ES 201 108 - Distributed speech recognition; Front-end feature extraction algorithm; Compression algorithms
Hinton, G.E.: A practical guide to training restricted Boltzmann machines. UTML TR 2010–003 (2010)
Theano Library. http://deeplearning.net/software/theano/
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2016 Springer International Publishing AG
About this paper
Cite this paper
López-Espejo, I., Peinado, A.M., Gomez, A.M., Martín-Doñas, J.M. (2016). Deep Neural Network-Based Noise Estimation for Robust ASR in Dual-Microphone Smartphones. In: Abad, A., et al. Advances in Speech and Language Technologies for Iberian Languages. IberSPEECH 2016. Lecture Notes in Computer Science(), vol 10077. Springer, Cham. https://doi.org/10.1007/978-3-319-49169-1_12
Download citation
DOI: https://doi.org/10.1007/978-3-319-49169-1_12
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-319-49168-4
Online ISBN: 978-3-319-49169-1
eBook Packages: Computer ScienceComputer Science (R0)