Learning local distance functions for place recognition

Result code in IS VaVaI
<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F68407700%3A21230%2F12%3A00200341" target="_blank" >RIV/68407700:21230/12:00200341 - isvavai.cz</a>
Result on the web
—
DOI - Digital Object Identifier
—

Result language
angličtina
Original language name
Learning local distance functions for place recognition
Original language description
The goal of this work is to retrieve images from a large geotagged image database depicting the same place as a given query photograph. Previous work has shown that objects, which frequently occur in the database (such as trees or road mark ings), can cause significant confusion between different places, and suppressing features on such "confusing" objects can significantly improve retrieval performance. In this work, we investigate whether suppressing confusing features in the database can be cast as learning local per-image distance functions. We demonstrate that finding such local distance functions can be formulated as a linear program and perform initial experiments on a database of 17,000 street view images of Paris.
Czech name
—
Czech description
—

Project
<a href="/en/project/7E10046" target="_blank" >7E10046: Planetary Robotics Vision Scout</a><br>
Continuities
P - Projekt vyzkumu a vyvoje financovany z verejnych zdroju (s odkazem do CEP)<br>S - Specificky vyzkum na vysokych skolach

Publication year
2012
Confidentiality
S - Úplné a pravdivé údaje o projektu nepodléhají ochraně podle zvláštních právních předpisů

Article name in the collection
CVWW 2012: Proceedings of the 17th Computer Vision Winter Workshop
ISBN
978-961-90901-6-9
ISSN
—
e-ISSN
—
Number of pages
6
Pages from-to
168-173
Publisher name
Slovenian Pattern Recognition Society
Place of publication
Ljubljana
Event location
Mala Nedelja
Event date
Feb 1, 2012
Type of event by nationality
WRD - Celosvětová akce
UT code for WoS article
—

Similar results(10)