Fast Spectral Ranking for Similarity Search - INRIA - Institut National de Recherche en Informatique et en Automatique Accéder directement au contenu
Pré-Publication, Document De Travail Année : 2017

Fast Spectral Ranking for Similarity Search

Résumé

Despite the success of deep learning on representing images for particular object retrieval, recent studies show that the learned representations still lie on manifolds in a high dimensional space. This makes the Euclidean nearest neighbor search biased for this task. Exploring the mani-folds online remains expensive even if a nearest neighbor graph has been computed offline. This work introduces an explicit embedding reducing manifold search to Euclidean search followed by dot product similarity search. This is equivalent to linear graph filtering of a sparse signal in the frequency domain. To speed up online search, we compute an approximate Fourier basis of the graph offline. We improve the state of art on particular object retrieval datasets including the challenging Instre dataset containing small objects. At a scale of 10 5 images, the offline cost is only a few hours, while query time is comparable to standard similarity search.
Fichier principal
Vignette du fichier
1703.06935.pdf (1.09 Mo) Télécharger le fichier
Origine : Fichiers produits par l'(les) auteur(s)
Loading...

Dates et versions

hal-02370336 , version 1 (19-11-2019)

Identifiants

Citer

Ahmet Iscen, Yannis Avrithis, Giorgos Tolias, Teddy Furon, Ondrej Chum. Fast Spectral Ranking for Similarity Search. 2017. ⟨hal-02370336⟩
44 Consultations
95 Téléchargements

Altmetric

Partager

Gmail Facebook X LinkedIn More