Abstract
Time-frequency constrained interpolation of audio has proven to be an effective technique in removing a wide variety of acoustic disturbances. Traditionally these techniques assume that the signal is stationary for the duration of the interpolation, which limits the types of disturbances that can be addressed. In this paper we propose masked positive semi-definite tensor factorisation followed by a novel form of multi-channel spectral subtraction to solve the problem, and we demonstrate excellent results on some real-world examples. The proposed methods can remove disturbances that were previously considered highly challenging to interpolate, for example a burst of wind noise in a voice recording.
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
Godsill, S., Rayner, P.: Digital Audio Restoration - A Statistical Model Based Approach, pp. 153–163. Springer, London (1998)
Betts, D.A.: Method and apparatus for audio signal processing. US Patent 7 978 862 (2011)
Fevotte, C., Bertin, N., Durrieu, J.-L.: Nonnegative matrix factorization with the Itakura-Saito Divergence: with application to music analysis. Neural Comput. 21(3), 793–830 (2008)
Févotte, C., Ozerov, A.: Notes on nonnegative tensor factorization of the spectrogram for audio source separation: statistical insights and towards self-clustering of the spatial cues. In: Aramaki, M., Jensen, K., Kronland-Martinet, R., Ystad, S. (eds.) CMMR 2010. LNCS, vol. 6684, pp. 102–115. Springer, Heidelberg (2011)
Ozerov, A., Fevotte, C.: Multichannel nonnegative matrix factorization in convolutive mixtures. With application to blind audio source separation. In: IEEE International Conference on Acoustics, Speech and Signal Processing. ICASSP 2009, pp. 3137–3140, April 2009
Sawada, H., Kameoka, H., Araki, S., Ueda, N.: Efficient algorithms for multichannel extensions of Itakura-Saito nonnegative matrix factorization. In: 2012 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), pp. 261–264, March 2012
Laurberg, H., Christensen, M.G., Plumbley, M.D., Hansen, L.K., Jensen, S.H.: Theorems on positive data: on the uniqueness of nmf. Comput. Intell. Neurosci. 2008, 1–9 (2008)
King, B., Févotte, C., Smaragdis, P.: Optimal cost function and magnitude power for NMF-based speech separation and music interpolation. In: 2012 IEEE International Workshop on Machine Learning for Signal Processing (MLSP), pp. 1–6. IEEE (2012)
Mohammadiha, N., Dodo, S.: Transient noise reduction using nonnegative matrix factorization. In: 2014 4th Joint Workshop on Hands-free Speech Communication and Microphone Arrays (HSCMA), pp. 27–31. IEEE (2014)
Bansal, D., Raj, B., Smaragdis, P.: Bandwidth expansion of narrowband speech using non-negative matrix factorization. In: INTERSPEECH, pp. 1505–1508 (2005)
Smaragdis, P., Raj, B., Shashanka, M.: Missing data imputation for spectral audio signals. In: IEEE International Workshop on Machine Learning for Signal Processing. MLSP 2009, pp. 1–6. IEEE (2009)
de Leeuw, J.: Block-relaxation algorithms in statistics. In: Bock, H.-H., Lenski, W., Richter, M.M. (eds.) Information Systems and Data Analysis. Studies in Classification, Data Analysis, and Knowledge Organization, pp. 308–324. Springer, Heidelberg (1994)
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2015 Springer International Publishing Switzerland
About this paper
Cite this paper
Betts, D. (2015). Masked Positive Semi-definite Tensor Interpolation. In: Vincent, E., Yeredor, A., Koldovský, Z., Tichavský, P. (eds) Latent Variable Analysis and Signal Separation. LVA/ICA 2015. Lecture Notes in Computer Science(), vol 9237. Springer, Cham. https://doi.org/10.1007/978-3-319-22482-4_52
Download citation
DOI: https://doi.org/10.1007/978-3-319-22482-4_52
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-319-22481-7
Online ISBN: 978-3-319-22482-4
eBook Packages: Computer ScienceComputer Science (R0)