Paper Info
Title
Underdetermined blind source separation based on Continuous Density Hidden Markov Models
Abstract
In this paper, a novel method is developed to solve the problem of underdetermined blind source separation, where the number of mixtures is smaller than that of sources. Generalized Gaussian Distributions (GGDs) are used to model the source signals and generative Continuous Density Hidden Markov Models (CDHMMs) are derived to track the nonstationarity inside the source signals. Each source signal can switch between several states such that the separation performance can be significantly improved. The model parameters are trained through the Expectation Maximization (EM) algorithm and the source signals are estimated via the Maximum a Posteriori (MAP) approach. Compared with the results of L1-norm solution, our proposed algorithm has obtained much better output signal-to-noise ratio (SNR) and the separation results are more realistic.
Year
DOI
Venue
2010
10.1109/ICASSP.2010.5495730
ICASSP
Keywords
Field
DocType
expectation-maximisation algorithm,expectation maximization algorithm and,source signal,underdetermined blind source separation,nonstationary signals,continuous density hidden markov model,generalized gaussian distributions,gaussian distribution,blind source separation,hidden markov model,maximum a posteriori approach,hidden markov models,generalized gaussian distribution,em algorithm,optimization,mathematical model,time frequency analysis,signal to noise ratio,switches,signal generators,null space,expectation maximization
Pattern recognition,Computer science,Expectation–maximization algorithm,Signal-to-noise ratio,Gaussian,Artificial intelligence,Maximum a posteriori estimation,Hidden Markov model,Blind signal separation,Source separation,Generalized normal distribution
Conference
ISSN
ISBN
Citations 
1520-6149 E-ISBN : 978-1-4244-4296-6
978-1-4244-4296-6
0
PageRank 
References 
Authors
0.34
3
2
Name
Order
Citations
PageRank
Xiaoming Zhu100.34
keshab k parhi23235369.07