In this paper we present and start analyzing the iCub World data-set, an object recognition data-set, we acquired using a Human-Robot Interaction (HRI) scheme and the iCub humanoid robot platform. Our set up allows for rapid acquisition and annotation of data with corresponding ground truth. While more constrained in its scopes -- the iCub world is essentially a robotics research lab -- we demonstrate how the proposed data-set poses challenges to current recognition systems. The iCubWorld data-set is publicly available. The data-set can be downloaded from: http://www.iit.it/en/projects/data-sets.html.
@article{arxiv.1306.3560,
title = {iCub World: Friendly Robots Help Building Good Vision Data-Sets},
author = {Sean Ryan Fanello and Carlo Ciliberto and Matteo Santoro and Lorenzo Natale and Giorgio Metta and Lorenzo Rosasco and Francesca Odone},
journal= {arXiv preprint arXiv:1306.3560},
year = {2013}
}
Comments
CVPR2013 Workshop: Ground Truth - What is a good dataset?. Portland, USA (June 28, 2013