@prefix dcat: . @prefix dct: . @prefix foaf: . @prefix vcard: . @prefix xsd: . a dcat:Dataset ; dct:description """ 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. """ ; dct:identifier "Report N°: RR-8209" ; dct:issued "2026-05-10T18:12:39.766748"^^xsd:dateTime ; dct:language "en" ; dct:modified "2026-05-10T18:12:39.766752"^^xsd:dateTime ; dct:publisher ; dct:title "Image Classification with the Fisher Vector: Theory and Practice" ; dcat:contactPoint [ a vcard:Organization ; vcard:fn "CCSD" ] ; dcat:distribution ; dcat:keyword "infoeu-reposemanticsreport", "infoinfo-lgcomputer-science-csmachine-learning-cslg", "reports" ; dcat:landingPage . a dcat:Distribution ; dct:format "HTML" ; dct:issued "2026-05-10T18:12:39.771603"^^xsd:dateTime ; dct:modified "2026-05-10T18:12:39.762000"^^xsd:dateTime ; dct:title "Image Classification with the Fisher Vector: Theory and Practice" ; dcat:accessURL . a foaf:Agent ; foaf:name "test_moissonnage_selune" . a foaf:Document .