A Dictionary Learning Method Based On EMD for Audio Sparse Representation

Sparse representation has long been studied and several
dictionary learning methods have been proposed. The dictionary
learning methods are widely used because they are adaptive. In this
paper, a new dictionary learning method for audio is proposed. Signals
are at first decomposed into different degrees of Intrinsic Mode
Functions (IMF) using Empirical Mode Decomposition (EMD)
technique. Then these IMFs form a learned dictionary. To reduce the
size of the dictionary, the K-means method is applied to the dictionary
to generate a K-EMD dictionary. Compared to K-SVD algorithm, the
K-EMD dictionary decomposes audio signals into structured
components, thus the sparsity of the representation is increased by
34.4% and the SNR of the recovered audio signals is increased by
20.9%.

• Yueming Wang
• Zenghui Zhang
• Rendong Ying
• Peilin Liu

• Dictionary Learning
• EMD
• K-means Method
• Sparse Representation.

<p>[1] B. Olshausen and D. Field, &ldquo;Emergence of simple-cell receptive
fieldproperties by learning a sparse code for natural images,&rdquo; Nature,
vol.381, pp. 607&ndash;609, 1996.
[2] M. S. Lewicki and T. J. Sejnowski, &ldquo;Learning overcomplete
representations,&rdquo; Neural Comput., vol. 12, pp. 337&ndash;365, 2000.
[3] P. Schmid-Saugeon and A. Zakhor, &ldquo;Dictionary design for
matchingpursuit and application to motion-compensated video coding,&rdquo;
IEEETrans. Circuits Syst. Video Technol., vol. 14, no. 6, pp. 880&ndash;886,
2004.
[4] M. Aharon, M. Elad, and A. Bruckstein, &ldquo;K-SVD: An algorithm for
designing overcomplete dictionaries for sparse representation,&rdquo;
IEEETrans. Signal Process, vol. 54, no. 11, pp. 4311&ndash;4322, 2006.
[5] M. Yaghoobi, L. Daudet, and M. Davies, &ldquo;Parametric dictionary design
for sparse coding,&rdquo; in Workshop on Signal Processing with Adaptive
Sparse Structured Representations (SPARS09), 2009
[6] M. G. Jafari and M. D. Plumbley. Fast dictionary learning for
sparserepresentations of speech signals. IEEE Journal of Selected Topics
in SignalProcessing, 5:1025&ndash;1031, Sep. 2011.
[7] N. E. Huang, Z. Shen, S. R. Long, M. C. Wu. H.H. Shih, et al, &ldquo;The
Empirical Mode Decompositionand The Hilbert Spectrum for Nonlinear
and Nonstationary Time Series Analysis&rdquo;, Proc. R. Soc. A,1998, 454, pp.
903-995.
[8] A. Gersho and R. M. Gray, Vector Quantization and Signal
Compression.Norwell, MA: Kluwer Academic, 1991.
[9] T. Blumensath and M. E. Davies &ldquo;Iterative thresholding for sparse
approximations&quot;, J. Fourier Anal. Applicat., vol. 14, no. 5, pp.629 -654
2008</p>