Efficient Diffusion on Region Manifolds: Recovering Small Objects with Compact CNN Representations

Kód výsledku v IS VaVaI

<a href="https://www.isvavai.cz/riv?ss=detail&h=RIV%2F68407700%3A21230%2F17%3A00312200" target="_blank" >RIV/68407700:21230/17:00312200 - isvavai.cz</a>

Výsledek na webu

<a href="http://dx.doi.org/10.1109/CVPR.2017.105" target="_blank" >http://dx.doi.org/10.1109/CVPR.2017.105</a>

DOI - Digital Object Identifier

<a href="http://dx.doi.org/10.1109/CVPR.2017.105" target="_blank" >10.1109/CVPR.2017.105</a>

Jazyk výsledku

angličtina

Název v původním jazyce

Efficient Diffusion on Region Manifolds: Recovering Small Objects with Compact CNN Representations

Popis výsledku v původním jazyce

Query expansion is a popular method to improve the quality of image retrieval with both conventional and CNN representations. It has been so far limited to global image similarity. This work focuses on diffusion, a mechanism that captures the image manifold in the feature space. The diffusion is carried out on descriptors of overlapping image regions rather than on a global image descriptor like in previous approaches. An efficient off-line stage allows optional reduction in the number of stored regions. In the on-line stage, the proposed handling of unseen queries in the indexing stage removes additional computation to adjust the precomputed data. We perform diffusion through a sparse linear system solver, yielding practical query times well below one second. Experimentally, we observe a significant boost in performance of image retrieval with compact CNN descriptors on standard benchmarks, especially when the query object covers only a small part of the image. Small objects have been a common failure case of CNN-based retrieval.

Název v anglickém jazyce

Efficient Diffusion on Region Manifolds: Recovering Small Objects with Compact CNN Representations

Popis výsledku anglicky

Query expansion is a popular method to improve the quality of image retrieval with both conventional and CNN representations. It has been so far limited to global image similarity. This work focuses on diffusion, a mechanism that captures the image manifold in the feature space. The diffusion is carried out on descriptors of overlapping image regions rather than on a global image descriptor like in previous approaches. An efficient off-line stage allows optional reduction in the number of stored regions. In the on-line stage, the proposed handling of unseen queries in the indexing stage removes additional computation to adjust the precomputed data. We perform diffusion through a sparse linear system solver, yielding practical query times well below one second. Experimentally, we observe a significant boost in performance of image retrieval with compact CNN descriptors on standard benchmarks, especially when the query object covers only a small part of the image. Small objects have been a common failure case of CNN-based retrieval.

Druh

D - Stať ve sborníku

CEP obor

—

OECD FORD obor

10201 - Computer sciences, information science, bioinformathics (hardware development to be 2.2, social aspect to be 5.8)

Projekt

<a href="/cs/project/LL1303" target="_blank" >LL1303: Vyhledávání vizuálních kategorií ve velkém množství obrázků</a><br>

Návaznosti

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

Rok uplatnění

2017

Kód důvěrnosti údajů

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

Název statě ve sborníku

CVPR 2017: Proceedings of the 2017 IEEE Conference on Computer Vision and Pattern Recognition

ISBN

978-1-5386-0457-1

ISSN

1063-6919

e-ISSN

—

Počet stran výsledku

10

Strana od-do

926-935

Název nakladatele

IEEE Computer Society Press

Místo vydání

—

Místo konání akce

Honolulu

Datum konání akce

21. 7. 2017

Typ akce podle státní příslušnosti

WRD - Celosvětová akce

Kód UT WoS článku

000418371400098