Opens in a new tab

AES E-Library

← Back to search

Convention Paper

Impulsive Noise Detection & Reduction using Adaptive Processing

Authors: Kalliris, George; Papaikolaou, George; Bahtsetzis, Stavros

AES Convention 96 · Paper 3870 · February 1994

Abstract

This paper describes new concepts on adaptive, impulsive noise detection & reduction algorithms using the well known autoregressive signal modeling theory and LMS based adaptive techniques. The concept of simultaneous detection and elimination of impulsive noises in the same process is first examined, resulting in simple but computationally expensive LPC based algorithm. In order to reduce computational cost, the separation of the detection and elimination processes, is then investigated yielding on a novel modified LMS algorithm which identifies the positions of the impulsive noises with much better performance over simple LMS. The elimination of the impulse by using various techniques is then described. The proposed technques are examined in various real - life audio recordings.

Details

Published in
AES Convention 96
AES Convention
96
Paper number
3870
Publication date
February 6, 1994
Session subject
Digital Signal Processing
Affiliation
Laboratory of Electroacoustics & Musical Technology, Department of Electrical Engineering, Aristotle Univerisity of Thessaloniki, Greece (See document for exact affiliation information.)
Type
Convention Paper