ROBIN Competition

Datasets, Ground Truths and Metrics
For the evaluation of
object recognition and image categorisation
algorithms

TESTS' DEFINITIONS

November 2006 : Added images and more details about the competitions ! Take a look at this document for more informations about the competitions. Take also a look the metrics definitions, and visit this page to know how to get the datasets.

The goal of these challenges is to recognize objects from a number of visual object classes in realistic scenes. It is fundamentally a supervised learning problem in that a training set of labelled images will be provided. The various object classes to research and the scenes observed depend on the datasets.

There will be two main competitions for each dataset :

- Some Classification Tasks

- Some Detection Tasks

For each of the 6 datasets, the tests might be slightly different. For more informations about each datasets test, please visit the links bellow.

Dataset #1 : Multi-class object detection with view point changes produced by Bertin Technologies and Cybernetix (ECA).

Dataset #2 : Generic Object classification in satellite images produced by CNES.

Dataset #3 : Multi-class object detection in Aerial images produced by SAGEM.

Dataset #4 : Object detection in aerial images produced by EADS.

Dataset #5 : Robustness of detection algorithms produced by MBDA.

Dataset #6 : Image categorisation produced by THALES.