Improving SIFT for Fast Tree Matching by Optimal Linear Projection

Result code in IS VaVaI

<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F68407700%3A21230%2F07%3A03135492" target="_blank" >RIV/68407700:21230/07:03135492 - isvavai.cz</a>

Result on the web

—

DOI - Digital Object Identifier

—

Result language

angličtina

Original language name

Improving SIFT for Fast Tree Matching by Optimal Linear Projection

Original language description

We propose to transform an image descriptor so that nearest neighbor (NN) search for correspondences becomes the optimal matching strategy under the assumption that inter-image deviations of corresponding descriptors have Gaussian distribution. The Euclidean NN in the transformed domain corresponds to the NN according to a truncated Mahalanobis metric in the original descriptor space. We provide theoretical justification for the proposed approach and show experimentally that the transformation allows asignificant dimensionality reduction and improves matching performance of a state-of-the art SIFT descriptor. We observe consistent improvement in precision-recall and speed of fast matching in tree structures at the expense of little overhead for projecting the descriptors into transformed space. In the context of SIFT vs. transformed MSIFT comparison, tree search structures are evaluated according to different criteria and query types. All search tree experiments confirm that transform

Czech name

Improving SIFT for Fast Tree Matching by Optimal Linear Projection

Czech description

We propose to transform an image descriptor so that nearest neighbor (NN) search for correspondences becomes the optimal matching strategy under the assumption that inter-image deviations of corresponding descriptors have Gaussian distribution. The Euclidean NN in the transformed domain corresponds to the NN according to a truncated Mahalanobis metric in the original descriptor space. We provide theoretical justification for the proposed approach and show experimentally that the transformation allows asignificant dimensionality reduction and improves matching performance of a state-of-the art SIFT descriptor. We observe consistent improvement in precision-recall and speed of fast matching in tree structures at the expense of little overhead for projecting the descriptors into transformed space. In the context of SIFT vs. transformed MSIFT comparison, tree search structures are evaluated according to different criteria and query types. All search tree experiments confirm that transform

Type

D - Article in proceedings

CEP classification

JD - Use of computers, robotics and its application

OECD FORD branch

—

Project

<a href="/en/project/GA102%2F07%2F1317" target="_blank" >GA102/07/1317: Methods for Visual Recognition of Large Collections of Non-rigid Objects</a><br>

Continuities

P - Projekt vyzkumu a vyvoje financovany z verejnych zdroju (s odkazem do CEP)

Publication year

2007

Confidentiality

S - Úplné a pravdivé údaje o projektu nepodléhají ochraně podle zvláštních právních předpisů

Article name in the collection

ICCV 2007: Proceedings of Eleventh IEEE International Conference on Computer Vision

ISBN

978-1-4244-1630-1

ISSN

—

e-ISSN

—

Number of pages

8

Pages from-to

—

Publisher name

Omnipress

Place of publication

Madison

Event location

Rio de Janeiro

Event date

Oct 14, 2007

Type of event by nationality

WRD - Celosvětová akce

UT code for WoS article

—