Neural filters: a class of filters unifying FIR and median filters

Author

Yin, Lin ; Astola, Jaakko ; Neuvo, Yrjö

Author_Institution

Dept. of Electr. Eng., Tampere Univ. of Technol., Finland

Volume

4

fYear

1992

fDate

23-26 Mar 1992

Firstpage

53

Abstract

A new class of nonlinear filters called neural filters based on the threshold decomposition and neural networks is introduced. Neural filters can approximate both linear finite impulse response (FIR) filters and weighted order statistic (WOS) filters which include median, rank order, and weighted median filters. An adaptive algorithm is derived for determining optimal neural filters under the mean squared error (MSE) criterion. Experimental results demonstrate that, if the input signal is corrupted by Gaussian noise, adaptive neural filters converge to linear filters and that, if corrupted by impulsive noise, optimal neural filters become WOS filters

Keywords

adaptive filters; digital filters; neural nets; Gaussian noise; impulsive noise; linear FIR filters; mean squared error; median filters; neural filters; neural networks; nonlinear filters; threshold decomposition; weighted order statistic filters; Adaptive filters; Adaptive systems; Finite impulse response filter; Gaussian noise; Linear approximation; Neural networks; Neurons; Noise cancellation; Nonlinear filters; Signal processing algorithms;

fLanguage

English

Publisher

ieee

Conference_Titel

Acoustics, Speech, and Signal Processing, 1992. ICASSP-92., 1992 IEEE International Conference on

Conference_Location

San Francisco, CA

ISSN

1520-6149

Print_ISBN

0-7803-0532-9

Type

conf

DOI

10.1109/ICASSP.1992.226413

Filename

226413