Code-excited neural vector quantization

Authors:
Zhicheng Wang;John V. Hanson
Affiliations:
Department of Electrical and Computer Engineering, University of Waterloo, Waterloo, Ontario, Canada;Department of Electrical and Computer Engineering, University of Waterloo, Waterloo, Ontario, Canada
Venue:
ICASSP'93 Proceedings of the 1993 IEEE international conference on Acoustics, speech, and signal processing: plenary, special, audio, underwater acoustics, VLSI, neural networks - Volume I
Year:
1993

Citing 1
Cited 0

On the approximate realization of continuous mappings by neural networks

Neural Networks

Quantified Score

Hi-index	0.00

Visualization

Abstract

The Generalized Lloyd Algorithm (GLA), better known as the Linde-Buzo-Gray (LBG) algorithm is the most widely used technique in classical vector quantization (VQ) for speech or image signal compression. However, the encoding complexity of the algorithm grows exponentially with the product of coding rate and vector dimension, which prohibits applying the technique to tasks with moderate to large encoding rates or vector dimensions. This paper presents a new VQ scheme which overcomes the successive search coding computation of traditional techniques by using a quasi-parallel mapping technique and makes VQ practical for higher encoding rates and/or vector dimensions. Neural computing techniques are used to implement parallel encoding and decoding mappings in VQ and the developed algorithm is applied to quantizing Gauss-Markov processes and artificial data sources. Comparisons of performance with the LBG algorithm are given.