Signal denoising using wavelet and block hidden Markov model
Liao Z.W.; Lam E.C.M.; Tang Y.Y.
Source PublicationInternational Conference on Machine Learning and Cybernetics
AbstractIn this paper, we propose a novel wavelet domain HMM using block to strike a delicate balance between improving spatial adaptability of contextual HMM (CHMM) and modeling a more reliable HMM. Each wavelet coefficient is modeled as a Guanssian Mixture model, and the dependencies among wavelet coefficients in each subband are described by a context structure, then the structure is modified by blocks which are connected areas in a scale conditioned on the same context. Before denoising signal, efficient Expectation Maximization (EM) algorithms are developed for fitting the HMMs to observational signal data. Parameters of trained HMM are used to modify wavelet coefficients according to the rule of minimizing the mean squared error (MSE) of signal. Then, reverse wavelet transformation is utilized to modify wavelet coefficients. Finally, experimental results are given. The results show Block hidden Markov model (BHMM) is a powerful yet simple tool in signal denoising.
KeywordBlock HMM Contextual HMM EM algorithm Hidden Markov Model (HMM)
URLView the original
Fulltext Access
Document TypeConference paper
CollectionUniversity of Macau
AffiliationHong Kong Baptist University
Recommended Citation
GB/T 7714
Liao Z.W.,Lam E.C.M.,Tang Y.Y.. Signal denoising using wavelet and block hidden Markov model[C],2003:2468-2471.
Files in This Item:
There are no files associated with this item.
Related Services
Recommend this item
Usage statistics
Export to Endnote
Google Scholar
Similar articles in Google Scholar
[Liao Z.W.]'s Articles
[Lam E.C.M.]'s Articles
[Tang Y.Y.]'s Articles
Baidu academic
Similar articles in Baidu academic
[Liao Z.W.]'s Articles
[Lam E.C.M.]'s Articles
[Tang Y.Y.]'s Articles
Bing Scholar
Similar articles in Bing Scholar
[Liao Z.W.]'s Articles
[Lam E.C.M.]'s Articles
[Tang Y.Y.]'s Articles
Terms of Use
No data!
Social Bookmark/Share
All comments (0)
No comment.

Items in the repository are protected by copyright, with all rights reserved, unless otherwise indicated.