Image Classification with the Fisher Vector: Theory and Practice

Jorge Sanchez 1, * Florent Perronnin 2, * Thomas Mensink 2, 3, * Jakob Verbeek 3, *
* Auteur correspondant
3 LEAR - Learning and recognition in vision
Inria Grenoble - Rhône-Alpes, LJK - Laboratoire Jean Kuntzmann, INPG - Institut National Polytechnique de Grenoble
Abstract : A standard approach to describe an image for classification and retrieval purposes is to extract a set of local patch descriptors, encode them into a high dimensional vector and pool them into an image-level signature. The most common patch encoding strategy consists in quantizing the local descriptors into a finite set of prototypical elements. This leads to the popular Bag-of-Visual words (BOV) representation. In this work, we propose to use the Fisher Kernel framework as an alternative patch encoding strategy: we describe patches by their deviation from an ''universal'' generative Gaussian mixture model. This representation, which we call Fisher Vector (FV) has many advantages: it is efficient to compute, it leads to excellent results even with efficient linear classifiers, and it can be compressed with a minimal loss of accuracy using product quantization. We report experimental results on five standard datasets -- PASCAL VOC 2007, Caltech 256, SUN 397, ILSVRC 2010 and ImageNet10K -- with up to 9M images and 10K classes, showing that the FV framework is a state-of-the-art patch encoding technique.
Type de document :
[Research Report] RR-8209, INRIA. 2013
Liste complète des métadonnées

Littérature citée [73 références]  Voir  Masquer  Télécharger
Contributeur : Jakob Verbeek <>
Soumis le : mercredi 12 juin 2013 - 15:40:49
Dernière modification le : mercredi 11 avril 2018 - 01:58:28
Document(s) archivé(s) le : mardi 4 avril 2017 - 20:52:17


Fichiers produits par l'(les) auteur(s)


  • HAL Id : hal-00779493, version 3



Jorge Sanchez, Florent Perronnin, Thomas Mensink, Jakob Verbeek. Image Classification with the Fisher Vector: Theory and Practice. [Research Report] RR-8209, INRIA. 2013. 〈hal-00779493v3〉



Consultations de la notice


Téléchargements de fichiers