Skip to Main content Skip to Navigation

Spatial Fisher Vectors for Image Categorization

Josip Krapac 1 Jakob Verbeek 1 Frédéric Jurie 1, 2
1 LEAR - Learning and recognition in vision
Inria Grenoble - Rhône-Alpes, LJK - Laboratoire Jean Kuntzmann, INPG - Institut National Polytechnique de Grenoble
2 Equipe Image - Laboratoire GREYC - UMR6072
GREYC - Groupe de Recherche en Informatique, Image, Automatique et Instrumentation de Caen
Abstract : We introduce an extension of bag-of-words image representations to encode spatial layout. Using the Fisher kernel framework we derive a representation that encodes the spatial mean and the variance of image regions associated with visual words. We extend this representation by using a Gaussian mixture model to encode spatial layout, and show that this model is related to a soft-assign version of the spatial pyramid representation. We also combine our representation of spatial layout with the use of Fisher kernels to encode the appearance of local features. Through an extensive experimental evaluation, we show that our representation yields state-of-the-art image categorization results, while being more compact than spatial pyramid representations. In particular, using Fisher kernels to encode both appearance and spatial layout results in an image representation that is computationally efficient, compact, and yields excellent performance while using linear classifiers.
Complete list of metadatas

Cited literature [31 references]  Display  Hide  Download
Contributor : Jakob Verbeek <>
Submitted on : Thursday, August 4, 2011 - 5:28:56 PM
Last modification on : Thursday, March 26, 2020 - 8:49:27 PM
Document(s) archivé(s) le : Monday, November 12, 2012 - 3:10:27 PM


Publisher files allowed on an open archive


  • HAL Id : inria-00613572, version 1


Josip Krapac, Jakob Verbeek, Frédéric Jurie. Spatial Fisher Vectors for Image Categorization. [Research Report] RR-7680, INRIA. 2011. ⟨inria-00613572⟩



Record views


Files downloads