Entropy-constrained vector quantizer design algorithm using competitive learning technique

Wen Jyi Hwang, Maw Rong Leou, Bo Yuan Ye, Shi Chiang Liao

Research output: Contribution to journalArticle

Abstract

A novel full-search variable-rate vector quantizer (VQ) design algorithm using competitive learning technique is presented. The algorithm, termed entropy-constrained competitive learning (ECCL) algorithm, can design a VQ having minimum average distortion subject to a rate constraint. The ECCL algorithm enjoys a better rate-distortion performance than that of the existing competitive learning algorithms. Moreover, the ECCL algorithm outperforms the entropy-constrained vector quantizer (ECVQ) design algorithm subject to the same rate and storage size constraints. In addition, the learning algorithm is more insensitive to the selection of initial codewords as compared with the ECVQ algorithm. Therefore, the ECCL algorithm can be an effective alternative to the existing variable-rate VQ design algorithms for the applications of signal compression.

Original languageEnglish
Pages (from-to)1715-1721
Number of pages7
JournalConference Record / IEEE Global Telecommunications Conference
Volume3
Publication statusPublished - 1998
Externally publishedYes

Fingerprint

Learning algorithms
entropy
Entropy
learning
compression
rate

ASJC Scopus subject areas

  • Electrical and Electronic Engineering
  • Global and Planetary Change

Cite this

Entropy-constrained vector quantizer design algorithm using competitive learning technique. / Hwang, Wen Jyi; Leou, Maw Rong; Ye, Bo Yuan; Liao, Shi Chiang.

In: Conference Record / IEEE Global Telecommunications Conference, Vol. 3, 1998, p. 1715-1721.

Research output: Contribution to journalArticle

@article{af46b7ce92e24066a9fb607fe9ede871,
title = "Entropy-constrained vector quantizer design algorithm using competitive learning technique",
abstract = "A novel full-search variable-rate vector quantizer (VQ) design algorithm using competitive learning technique is presented. The algorithm, termed entropy-constrained competitive learning (ECCL) algorithm, can design a VQ having minimum average distortion subject to a rate constraint. The ECCL algorithm enjoys a better rate-distortion performance than that of the existing competitive learning algorithms. Moreover, the ECCL algorithm outperforms the entropy-constrained vector quantizer (ECVQ) design algorithm subject to the same rate and storage size constraints. In addition, the learning algorithm is more insensitive to the selection of initial codewords as compared with the ECVQ algorithm. Therefore, the ECCL algorithm can be an effective alternative to the existing variable-rate VQ design algorithms for the applications of signal compression.",
author = "Hwang, {Wen Jyi} and Leou, {Maw Rong} and Ye, {Bo Yuan} and Liao, {Shi Chiang}",
year = "1998",
language = "English",
volume = "3",
pages = "1715--1721",
journal = "Conference Record / IEEE Global Telecommunications Conference",

}

TY - JOUR

T1 - Entropy-constrained vector quantizer design algorithm using competitive learning technique

AU - Hwang, Wen Jyi

AU - Leou, Maw Rong

AU - Ye, Bo Yuan

AU - Liao, Shi Chiang

PY - 1998

Y1 - 1998

N2 - A novel full-search variable-rate vector quantizer (VQ) design algorithm using competitive learning technique is presented. The algorithm, termed entropy-constrained competitive learning (ECCL) algorithm, can design a VQ having minimum average distortion subject to a rate constraint. The ECCL algorithm enjoys a better rate-distortion performance than that of the existing competitive learning algorithms. Moreover, the ECCL algorithm outperforms the entropy-constrained vector quantizer (ECVQ) design algorithm subject to the same rate and storage size constraints. In addition, the learning algorithm is more insensitive to the selection of initial codewords as compared with the ECVQ algorithm. Therefore, the ECCL algorithm can be an effective alternative to the existing variable-rate VQ design algorithms for the applications of signal compression.

AB - A novel full-search variable-rate vector quantizer (VQ) design algorithm using competitive learning technique is presented. The algorithm, termed entropy-constrained competitive learning (ECCL) algorithm, can design a VQ having minimum average distortion subject to a rate constraint. The ECCL algorithm enjoys a better rate-distortion performance than that of the existing competitive learning algorithms. Moreover, the ECCL algorithm outperforms the entropy-constrained vector quantizer (ECVQ) design algorithm subject to the same rate and storage size constraints. In addition, the learning algorithm is more insensitive to the selection of initial codewords as compared with the ECVQ algorithm. Therefore, the ECCL algorithm can be an effective alternative to the existing variable-rate VQ design algorithms for the applications of signal compression.

UR - http://www.scopus.com/inward/record.url?scp=0032260006&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0032260006&partnerID=8YFLogxK

M3 - Article

VL - 3

SP - 1715

EP - 1721

JO - Conference Record / IEEE Global Telecommunications Conference

JF - Conference Record / IEEE Global Telecommunications Conference

ER -