inria-00394211, version 3
On the burstiness of visual elements
Hervé Jégou
a, 1, 2Matthijs Douze a, 1Cordelia Schmid a, 1
IEEE Conference on Computer Vision and Pattern Recognition (CVPR '09) (2009) 1169-1176
Abstract: Burstiness, a phenomenon initially observed in text retrieval, is the property that a given visual element appears more times in an image than a statistically independent model would predict. In the context of image search, burstiness corrupts the visual similarity measure, i.e., the scores used to rank the images. In this paper, we propose a strategy to handle visual bursts for bag-of-features based image search systems. Experimental results on three reference datasets show that our method significantly and consistently outperforms the state of the art.
- a – INRIA
- 1: LEAR (INRIA Grenoble Rhône-Alpes / LJK Laboratoire Jean Kuntzmann)
- CNRS : FR71 – CNRS : UMR5527 – INRIA – Laboratoire Jean Kuntzmann – Université Joseph Fourier - Grenoble I – Institut National Polytechnique de Grenoble (INPG)
- 2: TEXMEX (INRIA - IRISA)
- CNRS : UMR6074 – INRIA – INSA Rennes – Université de Rennes 1
- Domain : Computer Science/Information Retrieval
Computer Science/Computer Vision and Pattern RecognitionComputer Science/Databases - Keywords : reference datasets – visual elements – burstiness – text retrieval – visual bursts – image search systems
- Available versions : v1 (2011-02-23) v2 (2011-06-07) v3 (2011-06-07)
- inria-00394211, version 3
- http://hal.inria.fr/inria-00394211
- oai:hal.inria.fr:inria-00394211
- From: Hervé Jégou
- Submitted on: Tuesday, 7 June 2011 10:24:50
- Updated on: Tuesday, 7 June 2011 10:32:46
Associated documents
Export