An efficient image similarity measure based on approximations of KL-divergence between two Gaussian mixtures

Jacob Goldberger, Shiri Gordon, Hayit Greenspan

Research output: Contribution to conferencePaperpeer-review

306 Scopus citations

Abstract

In this work we present two new methods for approximating the Kullback-Liebler (KL) divergence between two mixtures of Gaussians. The first method is based on matching between the Gaussian elements of the two Gaussian mixture densities. The second method is based on the unscented transform. The proposed methods are utilized for image retrieval tasks. Continuous probabilistic image modeling based on mixtures of Gaussians together with KL measure for image similarity, can be used for image retrieval tasks with remarkable performance. The efficiency and the performance of the KL approximation methods proposed are demonstrated on both simulated data and real image data sets. The experimental results indicate that our proposed approximations outperform previously suggested methods.

Original languageEnglish
Pages487-493
Number of pages7
DOIs
StatePublished - 1 Jan 2003
Externally publishedYes
EventProceedings: Ninth IEEE International Conference on Computer Vision - Nice, France
Duration: 13 Oct 200316 Oct 2003

Conference

ConferenceProceedings: Ninth IEEE International Conference on Computer Vision
Country/TerritoryFrance
CityNice
Period13/10/0316/10/03

ASJC Scopus subject areas

  • Software
  • Computer Vision and Pattern Recognition

Fingerprint

Dive into the research topics of 'An efficient image similarity measure based on approximations of KL-divergence between two Gaussian mixtures'. Together they form a unique fingerprint.

Cite this