Dissimilarity Computation through Low Rank Corrections

Dorin Comaniciu(1) Peter Meer(2) David Tyler(3)

(1)Imaging Research Department
Siemens Corporate Research
Princeton, NJ 08540

(2)Department of Electrical and Computer Engineering
(3)Department of Statistics
Rutgers University, Piscataway, NJ 08855, USA

Most of the energy of a multivariate feature is often contained in a low dimensional subspace. We exploit this property for the efficient computation of a dissimilarity measure between features using an approximation of the Bhattacharyya distance. We show that for normally distributed features the Bhattacharyya distance is a particular case of the Jensen-Shannon divergence, and thus evaluation of this distance is equivalent to a statistical test about the similarity of the two populations. The accuracy of the proposed approximation is tested for the task of texture retrieval.

Appeared in Pattern Recognition Letters, 24, 227-236, 2003.
Earlier version appeared in IEEE Workshop on Content-based Access of Image and Video Libraries (CBAIVL-99), Fort Collins, CO, June 1999, 50-54.

Return to Research: Content-based retrieval       Return to List of Publications
Download the paper