Optimal transform coding of Gaussian mixtures for joint classification/reconstruction

Soumya Jana, Pierre Moulin

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

In a variety of applications, classification systems operate on compressed signals. The design of optimal transform coders optimizing a joint classification/reconstruction criterion was explored, where classification accuracy is measured using the Chernoff bound on probability of misclassification and reconstruction quality is measured using mean-squared error (MSE) distortion. Under a high-rate assumption, local optimality properties of the Karhunen-Loeve transform (KLT) for a certain class of Gaussian mixtures under a joint classification/MSE measure was shown. Analytical expressions for optimal bit-allocation were derived. This generalizes classical optimality properties of the KLT for Gaussian sources under the MSE criterion.

Original languageEnglish (US)
Title of host publicationProceedings - DCC 2003
Subtitle of host publicationData Compression Conference
EditorsJames A. Storer, Martin Cohn
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages313-322
Number of pages10
ISBN (Electronic)0769518966
DOIs
StatePublished - 2003
EventData Compression Conference, DCC 2003 - Snowbird, United States
Duration: Mar 25 2003Mar 27 2003

Publication series

NameData Compression Conference Proceedings
Volume2003-January
ISSN (Print)1068-0314

Other

OtherData Compression Conference, DCC 2003
Country/TerritoryUnited States
CitySnowbird
Period3/25/033/27/03

Keywords

  • Bit rate
  • Compression algorithms
  • Data compression
  • Decorrelation
  • Design optimization
  • Distortion measurement
  • Image coding
  • Image reconstruction
  • Karhunen-Loeve transforms
  • Transform coding

ASJC Scopus subject areas

  • Computer Networks and Communications

Fingerprint

Dive into the research topics of 'Optimal transform coding of Gaussian mixtures for joint classification/reconstruction'. Together they form a unique fingerprint.

Cite this