Number of sources uncertainty in blind source separation: application to EMG signal processing
MetadataShow full item record
AbstractThis contribution deals with the number of components uncertainty in blind source separation. The number of components is estimated by maximizing its marginal a posteriori probability which favors the simplest explanation of the observed data. Marginalizing (integrating over all the parameters) is implemented through the Laplace approximation based on an efficient wavelet spectral matching separating algorithm. The effectiveness of the proposed method is shown on EMG data processing.
CitationSnoussi H, Khanna S, Hewson D, Duchene J (2007) 'Number of sources uncertainty in blind source separation: application to EMG signal processing', 29th Annual International Conference of the IEEE Engineering in Medicine and Biology Society - Lyon, IEEE.
JournalPROCEEDINGS OF THE 23RD ANNUAL INTERNATIONAL CONFERENCE OF THE IEEE ENGINEERING IN MEDICINE AND BIOLOGY SOCIETY, VOLS 1-4
TypeConference papers, meetings and proceedings
- Optimized wavelets for blind separation of nonstationary surface myoelectric signals.
- Authors: Farina D, Lucas MF, Doncarli C
- Issue date: 2008 Jan
- A new wavelet-based algorithm for compression of EMG signals.
- Authors: de A Berger P, de O Nascimento FA, da Rocha AF, Carvalho JL
- Issue date: 2007
- Covariance and time-scale methods for blind separation of delayed sources.
- Authors: Jiang N, Farina D
- Issue date: 2011 Mar
- Three-dimensional multiple-source focalization in an uncertain ocean environment.
- Authors: Tollefsen D, Dosso SE
- Issue date: 2013 Nov
- Unsupervised Bayesian decomposition of multiunit EMG recordings using Tabu search.
- Authors: Ge D, Le Carpentier E, Farina D
- Issue date: 2010 Mar