University of Surrey

Test tubes in the lab Research in the ATI Dance Research

BOLD - Binary online learned descriptor for efficient image matching.

Balntas, V, Tang, L and Mikolajczyk, K (2015) BOLD - Binary online learned descriptor for efficient image matching. In: CVPR 2015, 2015-06-07 - 2015-06-12, Boston, Mass..

[img]
Preview
Text
BOLD_01.pdf - Version of Record
Available under License : See the attached licence file.

Download (820kB) | Preview
[img]
Preview
PDF (licence)
SRI_deposit_agreement.pdf
Available under License : See the attached licence file.

Download (33kB) | Preview

Abstract

In this paper we propose a novel approach to generate a binary descriptor optimized for each image patch independently. The approach is inspired by the linear discriminant embedding that simultaneously increases inter and decreases intra class distances. A set of discriminative and uncorrelated binary tests is established from all possible tests in an offline training process. The patch adapted descriptors are then efficiently built online from a subset of tests which lead to lower intra class distances thus a more robust descriptor. A patch descriptor consists of two binary strings where one represents the results of the tests and the other indicates the subset of the patch-related robust tests that are used for calculating a masked Hamming distance. Our experiments on three different benchmarks demonstrate improvements in matching performance, and illustrate that per-patch optimization outperforms global optimization

Item Type: Conference or Workshop Item (Conference Paper)
Subjects : Computing Science
Divisions : Faculty of Engineering and Physical Sciences > Computing Science
Authors :
AuthorsEmailORCID
Balntas, VUNSPECIFIEDUNSPECIFIED
Tang, LUNSPECIFIEDUNSPECIFIED
Mikolajczyk, KUNSPECIFIEDUNSPECIFIED
Date : 2015
Identification Number : 10.1109/CVPR.2015.7298850
Copyright Disclaimer : This CVPR 2015 paper is the Open Access versions, provided by the Computer Vision Foundation. The authoritative version of this paper is posted on IEEE Xplore.
Contributors :
ContributionNameEmailORCID
UNSPECIFIEDIEEE, UNSPECIFIEDUNSPECIFIED
Related URLs :
Depositing User : Symplectic Elements
Date Deposited : 26 Jul 2016 10:09
Last Modified : 26 Jul 2016 10:09
URI: http://epubs.surrey.ac.uk/id/eprint/811371

Actions (login required)

View Item View Item

Downloads

Downloads per month over past year


Information about this web site

© The University of Surrey, Guildford, Surrey, GU2 7XH, United Kingdom.
+44 (0)1483 300800