Publication detail

Locally adaptive learning algorithm for separation of mixed signals

EKSLER, V., MALENOVSKÝ, M.

Original Title

Locally adaptive learning algorithm for separation of mixed signals

Type

conference paper

Language

English

Original Abstract

In this paper we want to introduce locally adaptive learning algorithm using feed-forward neural network for independent component analysis. The goal is to reconstruct original signals, which are mixed and picked up by the sensor array using blind signal processing techniques. In short we can formulate the problem as separation or estimation of the waveforms of the original sources from an array of sensors without knowing the characteristics of the transmission channel. We show a simulation of 5-input mixing system with ill-conditioned problems.

Keywords

source separation, mixed signals, blind signal processing, independent component analysis

Authors

EKSLER, V., MALENOVSKÝ, M.

RIV year

2004

Released

15. 9. 2004

Publisher

ČVUT

Location

Praha

ISBN

80-01-03063-6

Book

RTT 2004 - communication with future

Pages from

1

Pages to

4

Pages count

4

BibTex

@inproceedings{BUT11446,
  author="Václav {Eksler} and Marek {Malenovský}",
  title="Locally adaptive learning algorithm for separation of mixed signals",
  booktitle="RTT 2004 - communication with future",
  year="2004",
  pages="4",
  publisher="ČVUT",
  address="Praha",
  isbn="80-01-03063-6"
}