Supervised feature extraction for tensor objects based on maximization of mutual information

Jukić, Ante; Filipović, Marko (2013) Supervised feature extraction for tensor objects based on maximization of mutual information. Pattern Recognition Letters, 34 (13). pp. 1476-1484. ISSN 0167-8655

| Request a personal copy from author

Abstract

Several supervised feature extraction methods for tensor objects have been proposed recently, with applications in recognition of objects, faces and handwritten digits. However, the existing methods usually use only second order statistics of the data, typically through calculation of the within- and between-class scatters. Here we propose a method for supervised feature extraction for tensor objects based on maximization of an approximation of mutual information. In this way we utilize information contained in the higher order statistics of the data. Several experiments show that the proposed method results in highly discriminative features.

Item Type:	Article
Uncontrolled Keywords:	dimensionality reduction ; tensor decomposition ; feature extraction ; mutual information
Subjects:	TECHNICAL SCIENCES > Computing
Divisions:	Division of Laser and Atomic Research and Development
Projects:	Project title Project leader Project code Project type Analiza višespektralih podataka [217905] Ivica Kopriva 098-0982903-2558 MZOS
Depositing User:	Marko Filipović
Date Deposited:	13 May 2015 09:09
URI:	http://fulir.irb.hr/id/eprint/1921
DOI:	10.1016/j.patrec.2013.05.018

Actions (login required)

View Item