Proceedings of the conference have been published by

Springer-Verlag in the

Lecture Notes in

Computer Science Series (Volume 5534).


Click on the image to go to the publisher’s web site for this volume.

Click here for Top of Page

Workshop Report: GbR 2009

Report prepared by:


Andrea Torsello (Italy)

Text Box: 7th IAPR TC-15 Workshop on 
Graph-based Representations in Pattern Recognition 

26-28 May 2009
Venice, Italy
Right Arrow: Next
Right Arrow: Previous



Andrea Torsello (Italy)

Francisco Escolano (Spain)

Luc Brun (France)

GbR is a biennial workshop, organized by the IAPR Technical Committee 15 (TC-15) on Graph-based Representations, which aims to encourage research works in Pattern Recognition and Image Analysis within the graph theory framework. In May 2009, the 7th edition of workshop was held in the Ca' Dolfin palace in Venice, Italy, and was sponsored by the IAPR.

GbR 2009 has been a very successful workshop and one of the biggest in its history. The workshop received 47 submissions from 18 countries and 5 continents from which the Program committee selected 18 for oral presentation and 19 as posters. The papers presented in the workshop covered the use of graphs at all levels of representation, from low level image segmentation to high level human behavior. There were papers on formalizing the use of graphs for representing and recognizing data ranging from visual shape to music, papers focusing on the development of new and efficient approaches to matching graphs, on the use of graphs for supervised and unsupervised classification, on learning the structure of sets of graphs, and on the use of graph pyramids and combinatorial maps to provide suitable coarse-to-fine representations. Encouragingly, the workshop saw the convergence of ideas from several fields, from Spectral Graph Theory to Machine Learning to Graphics.

The workshop included two keynote addresses. The fist keynote speech was presented by Amnon Shashua (The Hebrew University of Jerusalem, Israel) who described a general scheme for message passing update rules for approximate inference in graphical models based on the framework of Fenchel duality. The second keynote speech was delivered by Ramin Zabih (Cornell University, USA) who gave an overview of the latest results concerning the use of graph cut algorithms for linear inverse systems.

The conference was run over 3 days, with very well populated oral sessions right through to the end of the conference. The poster sessions were bustling and were an excellent forum for discussion and feedback.

The proceedings of the conference have been published in the Lecture Notes in Computer Science series of Springer.