论文信息 - A METHOD OF ICA IN TIME-FREQUENCY DOMAIN

A METHOD OF ICA IN TIME-FREQUENCY DOMAIN

AMETHODOFICAINTIME-FREQUENCYDOMAINShiroIkedaPRESTO,JSTHirosawa2-1,Wako,351-0198,JapanShiro.Ikeda@brain.riken.go.jpNoboruMurataRIKENBSINob oru.Murata@brain.rikABSTRACTWeprop oseametho dofICAforseparatingconvolu-tivemixturesofacousticsignals.Thesignalsrecordedinarealenvironmentarenotinstantaneousbutconvolutivemixtures,b ecauseofthedelayandreections.Inordertoseparatethesesignals,itisef-fectivetotransformthesignalsintime-frequencydo-main.Thedicultp ointintheseapproachesistheambiguityofthep ermutationandamplitudewhichisunavoidableinoriginalICAproblem.Sincetheba-sicICAapproachescannotsolvetheseambiguity,wneedanotherapproachtosolvethem.Wemployedtheenvelop esofsignalstosolveit,andhade-velop edsomealgorithms.Inthisarticle,weshotheoutlineofourorignalmetho d,andsomeextensionsit.Theyare,theon-lineversionandauditorysceneanalysisproblem.1.INTRODUCTIONOneofthego o dapplicationsICAisseparatingacousticsignalsrecordedinarealenvironment.Thisproblemiswellknownasthename,\co cktailpartyproblem".Whatisdicultfortheseproblemsthatthesignalsincludedelaysandreections(Fig.1).

Shiro Ikeda | Noboru Murata | N. Murata | Shiro Ikeda

[1] Paris Smaragdis,et al. Blind separation of convolved mixtures in the frequency domain , 1998, Neurocomputing.

[2] Andreas Ziehe,et al. Combining time-delayed decorrelation and ICA: towards solving the cocktail party problem , 1998, Proceedings of the 1998 IEEE International Conference on Acoustics, Speech and Signal Processing, ICASSP '98 (Cat. No.98CH36181).

[3] Noboru Murata,et al. An Approach to Blind Source Separation of Speech Signals , 1998 .

[4] N. Murata,et al. An On-line Algorithm for Blind Source Separation on Speech Signals , 1998 .

[5] Andreas Ziehe,et al. An approach to blind source separation based on temporal structure of speech signals , 2001, Neurocomputing.

[6] Andrzej Cichocki,et al. Stability Analysis of Learning Algorithms for Blind Source Separation , 1997, Neural Networks.

[7] Terrence J. Sejnowski,et al. An Information-Maximization Approach to Blind Separation and Blind Deconvolution , 1995, Neural Computation.

[8] Schuster,et al. Separation of a mixture of independent signals using time delayed correlations. , 1994, Physical review letters.

[9] Andrzej Cichocki,et al. Neural networks for blind decorrelation of signals , 1997, IEEE Trans. Signal Process..