The Eff2 Image Retrieval System Prototype

Ragnheiður Grétarsdóttir (1) , Sigurðir Einarsson (1) , Björn Þór Jónsson (1) , Laurent Amsaleg (2)

1 School of Computer Science [Reykjavik]
2 TEXMEX - Multimedia content-based indexing

In recent years, content-based image retrieval has become more and more important in many fields. In the Eff2 project we have been working with one of the most advanced image description schemes available, namely a fine-grained image recognition scheme based on local descriptors. While this description scheme has been used successfully in the demanding application of content-based image copy right protection, query processing in this approach is very slow. This paper describes the Eff2 image retrieval system prototype, which is designed to facilitate user interaction with the local descriptor search process. We demonstrate, through experimental results, that the time to obtain a good answer may be reduced significantly, in some cases by over 85%, by allowing the user to examine intermediate results.

Domaines: Base de données [cs.DB]; Multimédia [cs.MM]

Ragnheiður Grétarsdóttir, Sigurðir Einarsson, Björn Þór Jónsson, Laurent Amsaleg. The Eff2 Image Retrieval System Prototype. IASTED International Conference on Databases and Applications, Feb 2005, Innsbruck, Austria. ⟨inria-00175405⟩