Abstract
We propose a novel unsupervised single-channel approach to separate between stationary and non-stationary signals. To this, we enhance data representation through its time-frequency space, where stationarity is defined based on information theory. Then, we search for the projection of the time-frequency representation that is as stationarity as possible, but preserving most of the data information. The proposed approach validated on synthetic data. As performance measure, we use the correlation coefficient and the mean squared error between the original and the estimated stationary composing signals. Obtained results are compared against the baseline non-negative matrix factorization that separates dynamics from the time-frequency representation. As a result, our approach gets better performance even if assuming low power ratios, i.e., non-stationary signal power is higher or even equal than the stationary signal power.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
References
Martínez-Vargas, J., Sepulveda-Cano, L., Travieso-Gonzalez, C., Castellanos-Dominguez, G.: Detection of obstructive sleep apnoea using dynamic filter-banked features. Expert Syst. Appl. 39(10), 9118–9128 (2012)
Hopgood, J., Rayner, P.J.W.: Single channel nonstationary stochastic signal separation using linear time-varying filters. IEEE Trans. Signal Process. 51(7), 1739–1752 (2003)
Gu, F., Zhang, H., Tan, X., Zhu, D.: Rls algorithm for blind source separation in non-stationary environments. In: Symposium on ICT and Energy Efficiency and Workshop on Information Theory and Security (CIICT 2012), pp. 162–165, July 2012
Mijovicc, B., De Vos, M., Gligorijevicc, I., Taelman, J., Van Huffel, S.: Source separation from single-channel recordings by combining empirical-mode decomposition and independent component analysis. IEEE Trans. Biomed. Eng. 57(9), 2188–2196 (2010)
Gao, B., Woo, W., Dlay, S.: Unsupervised single-channel separation of nonstationary signals using gammatone filterbank and itakura-saito nonnegative matrix two-dimensional factorizations. IEEE Trans. Circ. Syst. I Regul. Pap. 60(3), 662–675 (2013)
Tong, S., Li, Z., Zhu, Y., Thakor, N.V.: Describing the nonstationarity level of neurological signals based on quantifications of time-frequency representation. IEEE Trans. Biomed. Eng. 54(10), 1780–1785 (2007)
Principe, J.C.: Information Theoretic Learning: Renyi’s Entropy and Kernel Perspectives, 1st edn. Springer Publishing Company, Incorporated, Berlin (2010)
Rao, S.M.: Unsupervised learning: an information theoretic framework. Ph.D. dissertation, Gainesville, FL, USA (2008). aAI3367034
Acknowledgments
This research is carried out under the grant project 22056 - DESARROLLO DE UN SISTEMA AUTOMÁTICO DE MAPEO CEREBRAL PARA EL MONITOREO Y TRATAMIENTO DE ENFERMEDADES NEURODEGENERATIVAS sponsored by DIMA from Universidad Nacional de Colombia - Manizales. It is also supported by the scholarship Programa Nacional de Formacion de Investigadores GENERACION DEL BICENTENARIO, 2011 and project Evaluación asistida de potenciales evocados cognitivos como marcador del transtorno por déficit de atensión e hiperactividad (TDAH) founded by COLCIENCIAS.
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
Martínez-Vargas, J.D., Castro-Hoyos, C., Espinosa-Oviedo, J.J., Álvarez-Mesa, A.M., Castellanos-Dominguez, G. (2015). Single-Channel Separation Between Stationary and Non-stationary Signals Using Relevant Information. In: Paredes, R., Cardoso, J., Pardo, X. (eds) Pattern Recognition and Image Analysis. IbPRIA 2015. Lecture Notes in Computer Science(), vol 9117. Springer, Cham. https://doi.org/10.1007/978-3-319-19390-8_51
Download citation
DOI: https://doi.org/10.1007/978-3-319-19390-8_51
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-319-19389-2
Online ISBN: 978-3-319-19390-8
eBook Packages: Computer ScienceComputer Science (R0)