NetVLAD: CNN architecture for weakly supervised place recognition

Kód výsledku v IS VaVaI

<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F68407700%3A21230%2F16%3A00304272" target="_blank" >RIV/68407700:21230/16:00304272 - isvavai.cz</a>

Výsledek na webu

<a href="http://80.ieeexplore.ieee.org.dialog.cvut.cz/document/7780941/" target="_blank" >http://80.ieeexplore.ieee.org.dialog.cvut.cz/document/7780941/</a>

DOI - Digital Object Identifier

—

Jazyk výsledku

angličtina

Název v původním jazyce

NetVLAD: CNN architecture for weakly supervised place recognition

Popis výsledku v původním jazyce

We tackle the problem of large scale visual place recognition, where the task is to quickly and accurately recognize the location of a given query photograph. We present the following three principal contributions. First, we develop a convolutional neural network (CNN) architecture that is trainable in an end-to-end manner directly for the place recognition task. The main component of this architecture, NetVLAD, is a new generalized VLAD layer, inspired by the "Vector of Locally Aggregated Descriptors" image representation commonly used in image retrieval. The layer is readily pluggable into any CNN architecture and amenable to training via backpropagation. Second, we develop a training procedure, based on a new weakly supervised ranking loss, to learn parameters of the architecture in an end-to-end manner from images depicting the same places over time downloaded from Google Street View Time Machine. Finally, we show that the proposed architecture significantly outperforms non-learnt image representations and off-the-shelf CNN descriptors on two challenging place recognition benchmarks, and improves over current stateofthe-art compact image representations on standard image retrieval benchmarks.

Název v anglickém jazyce

NetVLAD: CNN architecture for weakly supervised place recognition

Popis výsledku anglicky

We tackle the problem of large scale visual place recognition, where the task is to quickly and accurately recognize the location of a given query photograph. We present the following three principal contributions. First, we develop a convolutional neural network (CNN) architecture that is trainable in an end-to-end manner directly for the place recognition task. The main component of this architecture, NetVLAD, is a new generalized VLAD layer, inspired by the "Vector of Locally Aggregated Descriptors" image representation commonly used in image retrieval. The layer is readily pluggable into any CNN architecture and amenable to training via backpropagation. Second, we develop a training procedure, based on a new weakly supervised ranking loss, to learn parameters of the architecture in an end-to-end manner from images depicting the same places over time downloaded from Google Street View Time Machine. Finally, we show that the proposed architecture significantly outperforms non-learnt image representations and off-the-shelf CNN descriptors on two challenging place recognition benchmarks, and improves over current stateofthe-art compact image representations on standard image retrieval benchmarks.

Druh

D - Stať ve sborníku

CEP obor

JD - Využití počítačů, robotika a její aplikace

OECD FORD obor

—

Projekt

<a href="/cs/project/7E13015" target="_blank" >7E13015: Planetary Robotics Data Exploitation</a><br>

Návaznosti

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

Rok uplatnění

2016

Kód důvěrnosti údajů

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

Název statě ve sborníku

CVPR 2016: Proceedings of the 2016 IEEE Conference on Computer Vision and Pattern Recognition

ISBN

978-1-4673-8851-1

ISSN

1063-6919

e-ISSN

—

Počet stran výsledku

11

Strana od-do

5297-5307

Název nakladatele

IEEE Computer Society Press

Místo vydání

Los Alamitos

Místo konání akce

Las Vegas

Datum konání akce

26. 6. 2016

Typ akce podle státní příslušnosti

WRD - Celosvětová akce

Kód UT WoS článku

—