I’ve presented the paper “Matching Local Descriptors for Image Identification on Cultural Databases” on a poster session of the 9th International Conference on Document Analysis and Recognition — ICDAR 2007. This paper was about the problem of image identification (or copy detection) and the 3-way tree, the first original method I have devised in my Ph.D. work for the indexing of high-dimensional multimedia descriptors.

Here’s the abstract: “In this paper we present a new method for high-dimensional descriptor matching, based on the KD-Tree, which is a classic method for nearest neighbours search. This new method, which we name 3-Way Tree, avoids the boundary effects that disrupt the KD-Tree in higher dimensionalities, by the addition of redundant, overlapping sub-trees. That way, more precision is ob-tained for the same querying times. We evaluate our method in the context of image identification for cul-tural collections, a task which can greatly benefit from the use of high-dimensional local descriptors computed around PoI (Points of Interest).”

Poster sessions can be either a hit or a miss, depending on how well the conference organisation integrates them to the other activities going around. Fortunately, in ICDAR 2007 the Posters were very next to the heart of the conference, the place where everything happens: the snacks table. Therefore the sessions were very interactive and interesting.

The fulltext of the paper and the poster can be found in my publications page.