Permutation Based Indexing for High Dimensional Data on GPU Architectures

Kód výsledku v IS VaVaI

<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F00216208%3A11320%2F15%3A10294742" target="_blank" >RIV/00216208:11320/15:10294742 - isvavai.cz</a>

Výsledek na webu

<a href="http://dx.doi.org/10.1109/CBMI.2015.7153619" target="_blank" >http://dx.doi.org/10.1109/CBMI.2015.7153619</a>

DOI - Digital Object Identifier

<a href="http://dx.doi.org/10.1109/CBMI.2015.7153619" target="_blank" >10.1109/CBMI.2015.7153619</a>

Jazyk výsledku

angličtina

Název v původním jazyce

Permutation Based Indexing for High Dimensional Data on GPU Architectures

Popis výsledku v původním jazyce

Permutation-based indexing is one of the most popular techniques for the approximate nearest-neighbor search problem in high-dimensional spaces. Due to the exponential increase of multimedia data, the time required to index this data has become a seriousconstraint of the indexing techniques. One of the possible steps towards faster index construction is utilization of massively parallel platforms such as the GPGPU architectures. In this paper, we have analyzed the computational costs of individual steps of the permutation-based index construction in a high-dimensional feature space and proposed a hybrid solution, where computational power of GPU is utilized for distance computations whilst the host CPU performs the postprocessing and sorting steps. Despite the fact that computing the distances is a naturally data-parallel task, an efficient implementation is quite challenging due to various GPU limitations and complex memory hierarchy. We have tested possible approaches to work divisi

Název v anglickém jazyce

Permutation Based Indexing for High Dimensional Data on GPU Architectures

Popis výsledku anglicky

Permutation-based indexing is one of the most popular techniques for the approximate nearest-neighbor search problem in high-dimensional spaces. Due to the exponential increase of multimedia data, the time required to index this data has become a seriousconstraint of the indexing techniques. One of the possible steps towards faster index construction is utilization of massively parallel platforms such as the GPGPU architectures. In this paper, we have analyzed the computational costs of individual steps of the permutation-based index construction in a high-dimensional feature space and proposed a hybrid solution, where computational power of GPU is utilized for distance computations whilst the host CPU performs the postprocessing and sorting steps. Despite the fact that computing the distances is a naturally data-parallel task, an efficient implementation is quite challenging due to various GPU limitations and complex memory hierarchy. We have tested possible approaches to work divisi

Druh

D - Stať ve sborníku

CEP obor

IN - Informatika

OECD FORD obor

—

Projekt

<a href="/cs/project/GP14-14292P" target="_blank" >GP14-14292P: Nasazení moderních paralelních architektur ve specifických oblastech databázových systémů</a><br>

Návaznosti

I - Institucionalni podpora na dlouhodoby koncepcni rozvoj vyzkumne organizace

Rok uplatnění

2015

Kód důvěrnosti údajů

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

Název statě ve sborníku

13th International Workshop on Content-Based Multimedia Indexing (CBMI)

ISBN

978-1-4673-6870-4

ISSN

1949-3991

e-ISSN

—

Počet stran výsledku

6

Strana od-do

1-6

Název nakladatele

IEEE

Místo vydání

neuveden

Místo konání akce

Prague, Czech Republic

Datum konání akce

10. 6. 2015

Typ akce podle státní příslušnosti

WRD - Celosvětová akce

Kód UT WoS článku

—