Presentation 1999/5/20
A Constraint Learning Algorithrn for Blind Source Separation
Motoki Nitta, Kenji Nakayama, Akihiro Hirano,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) A constrained learning algorithm is proposed to stabilize the Jutten's algorithm. The output samples with large amplitude are detected based on the standard deviation, and the learning is skipped for these output samples. Imbalance of the source signal powers is estimated by the correlation coefficient .of the observed signals. When the correlation coefficient is high, the correction term by the Jutten's algorithm becomes wrong value. Therefore, adjusting the parameters is skipped. Computer simulation using many kinds of signal sources have been carried out. The results demonstrate the proposed constrained learning algorithm works well for the signal sources with asymmetrical PDFs and imbalanced sigpal powers.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) Blind separation / Learning / Stability / Probability density function / Correlation coefficient / Speech
Paper # NC99-9
Date of Issue

Conference Information
Committee NC
Conference Date 1999/5/20(1days)
Place (in Japanese) (See Japanese page)
Place (in English)
Topics (in Japanese) (See Japanese page)
Topics (in English)
Chair
Vice Chair
Secretary
Assistant

Paper Information
Registration To Neurocomputing (NC)
Language JPN
Title (in Japanese) (See Japanese page)
Sub Title (in Japanese) (See Japanese page)
Title (in English) A Constraint Learning Algorithrn for Blind Source Separation
Sub Title (in English)
Keyword(1) Blind separation
Keyword(2) Learning
Keyword(3) Stability
Keyword(4) Probability density function
Keyword(5) Correlation coefficient
Keyword(6) Speech
1st Author's Name Motoki Nitta
1st Author's Affiliation Div. of Electronics and Computer Science, Master's Level Section Graduate School of Natural Science and Technology, Kanazawa Univ.:Matsushita Electric Industrial Co.()
2nd Author's Name Kenji Nakayama
2nd Author's Affiliation Div. of Electronics and Computer Science, Master's Level Section Graduate School of Natural Science and Technology, Kanazawa Univ.
3rd Author's Name Akihiro Hirano
3rd Author's Affiliation Div. of Electronics and Computer Science, Master's Level Section Graduate School of Natural Science and Technology, Kanazawa Univ.
Date 1999/5/20
Paper # NC99-9
Volume (vol) vol.99
Number (no) 58
Page pp.pp.-
#Pages 8
Date of Issue