Skip to Main content Skip to Navigation
New interface
Journal articles

Graph-based Particular Object Discovery

Oriane Siméoni 1 Ahmet Iscen 1, * Giorgos Tolias 2 Yannis Avrithis 1 Ondřej Chum 2 
* Corresponding author
1 LinkMedia - Creating and exploiting explicit links between multimedia fragments
Inria Rennes – Bretagne Atlantique , IRISA-D6 - MEDIA ET INTERACTIONS
2 VRG - Visual Recognition Group [Prague]
FEL CTU - Faculty of Electrical Engineering [Prague]
Abstract : Severe background clutter is challenging in many computer vision tasks, including large-scale image retrieval. Global descriptors, that are popular due to their memory and search efficiency, are especially prone to corruption by such a clutter. Eliminating the impact of the clutter on the image descriptor increases the chance of retrieving relevant images and prevents topic drift due to actually retrieving the clutter in the case of query expansion. In this work, we propose a novel salient region detection method. It captures, in an unsupervised manner, patterns that are both discriminative and common in the dataset. Saliency is based on a central-ity measure of a nearest neighbor graph constructed from regional CNN representations of dataset images. The proposed method exploits recent CNN architectures trained for object retrieval to construct the image representation from the salient regions. We improve particular object retrieval on challenging datasets containing small objects.
Document type :
Journal articles
Complete list of metadata

Cited literature [58 references]  Display  Hide  Download
Contributor : Yannis Avrithis Connect in order to contact the contributor
Submitted on : Tuesday, November 19, 2019 - 1:04:44 PM
Last modification on : Friday, August 5, 2022 - 2:54:52 PM
Long-term archiving on: : Thursday, February 20, 2020 - 4:47:44 PM


Files produced by the author(s)



Oriane Siméoni, Ahmet Iscen, Giorgos Tolias, Yannis Avrithis, Ondřej Chum. Graph-based Particular Object Discovery. Machine Vision and Applications, 2019, 30 (2), pp.243-254. ⟨10.1007/s00138-019-01005-z⟩. ⟨hal-02370238⟩



Record views


Files downloads