Overall Objectives
Scientific Foundations
Application Domains
New Results
Contracts and Grants with Industry
Other Grants and Activities

Section: Other Grants and Activities

International Collaborations

Collaboration with Reykjavík University, Iceland

Keywords : Content-Based Image Retrieval Systems, Curse of Dimensionality, Approximate Search Schemes, Random Projections, Local Descriptors, User Interface.

Participants : Laurent Amsaleg, Hervé Jégou.

This collaboration is done in the context of the INRIA Associate Teams program. This program links two research teams (one INRIA, one foreign) willing to cross-leverage their respective excellence and their complementarity. Björn Þór-Jónsson (Associate Professor) leads the team of researchers involved in Iceland.

Image databases, and content-based image retrieval systems in particular, have become increasingly important in many applications areas. Moreover, new applications exploiting fine detail of images are now fast emerging thanks to recent and modern image processing techniques. While extremely effective (they return high quality results), these image processing techniques are very inefficient (they answer very slowly) due to their complexity and because of the inadequacy of traditional lower layers of software. This is particularly prevalent at large scale when dealing with image collections of realistic sizes. The goal of this project is to research and develop new database support that integrates efficiency and effectiveness for modern large-scale computer-vision related applications and problems.

Together, we came up with the PvS-framework that provides an efficient and scalable support for local description based recognition applications. While this work is still very active, we have initiated another thread of research by investigating the browsing of personnal image collections. Today, everyone can witness the tremendous increase in the capability to create, share and store digital images. As a result, personal image collections are growing at an astounding rate and it is clear that in the future individuals will need to access tens of thousands, or even hundreds of thousands, of digital images. It is therefore imperative to start studying ways to access these images in a useful and interesting manner. Addressing this topic is a new development in our cooperation.

Collaboration with Croatia and Slovenia

Participant : Annie Morin.

Partners: Andrija Stampar School of Public Health, Medical School, University of Zagreb, department of Electronics, Microelectronics, Computer and Intelligent systems, University of Zagreb, Zagreb, Croatia; Faculty of Computer and Information Science, University of Ljubljana, Slovenia; ERIC lab., University of Lyon; Rudjer Boskovic Institute, Zagreb, Croatia.

A. Morin keeps on working with the university of Zagreb and the university of Ljubljana in spite of the end of the Egide cooperation program. She was invited in April to give a course to Master students at the university of Zagreb. One of the Master student wrote his thesis on the use of factorial correspondence analysis in text mining and its implementation in the free software Orange, a data mining library for Python developed by Blaz Zupan at the university of Ljubljana.

Collaboration with Dublin City University, Ireland

Participants : Laurent Amsaleg, Patrick Gros.

Our collaboration in the frame of aceMedia outlined the opportunity for a deeper collaboration on large image base indexing and retrieval. The first objective of this new collaboration is to make a first experience of indexing and retrieval with a database of one million images. DCU has a good knowledge on image descriptors and retrieval system interfaces which is complementary to ours. On the other hand, our indexing technology is more advanced. A corpus of several hundreds of thousands of images has been shared between DCU and TexMex to make common experiments.

Collaboration with University of Montreal, Quebec, Canada

Participants : Vincent Claveau, Stéphane Huet, Fabienne Moreau, Pascale Sébillot.

A one-year collaboration agreement has been signed between Montreal (OLST research project) and Rennes 1 (TexMex research project) universities mid-2005. The cooperation aims at automatically detecting in textual corpora paradigmatic relations between lexical units or terms.


Logo Inria