DocumentCode
2997469
Title
Post-nonlinear source separation: hard switching versus soft learning
Author
Chen, Yang ; He, Zhenya
Author_Institution
Dept. of Radio Eng., Southeast Univ., Nanjing, China
fYear
2000
fDate
2000
Firstpage
403
Lastpage
406
Abstract
Post-nonlinear mixtures give a practical nonlinear mixing scenario. Multilayer perceptron is a good choice for adjusting the post-nonlinearity and is taken in both of the two given post-nonlinear source separation algorithms. The difference lies in that the first one switches between fixed distribution models while the second realizes a soft learning on a new flexible yet simple distribution model
Keywords
adaptive signal processing; learning (artificial intelligence); multilayer perceptrons; MLP; blind source separation; fixed distribution models; flexible distribution model; hard switching; multilayer perceptron; nonlinear mixing scenario; post-nonlinear source separation; post-nonlinearity adjustment; soft learning; source separation algorithms; Algorithm design and analysis; Digital signal processing; Helium; Laboratories; Maximum likelihood estimation; Multilayer perceptrons; Nonlinear distortion; Source separation; Speech; Switches;
fLanguage
English
Publisher
ieee
Conference_Titel
Circuits and Systems, 2000. IEEE APCCAS 2000. The 2000 IEEE Asia-Pacific Conference on
Conference_Location
Tianjin
Print_ISBN
0-7803-6253-5
Type
conf
DOI
10.1109/APCCAS.2000.913520
Filename
913520
Link To Document