Accéder directement au contenu Accéder directement à la navigation
Article dans une revue

Virtual Training for a Real Application: Accurate Object-Robot Relative Localization without Calibration

Abstract : Localizing an object accurately with respect to a robot is a key step for autonomous robotic manipulation. In this work, we propose to tackle this task knowing only 3D models of the robot and object in the particular case where the scene is viewed from uncalibrated cameras — a situation which would be typical in an uncontrolled environment, e.g., on a construction site. We demonstrate that this localization can be performed very accurately, with millimetric errors, without using a single real image for training, a strong advantage since acquiring representative training data is a long and expensive process. Our approach relies on a classification Convolutional Neural Network (CNN) trained using hundreds of thousands of synthetically rendered scenes with randomized parameters. To evaluate our approach quantitatively and make it comparable to alternative approaches, we build a new rich dataset of real robot images with accurately localized blocks.
Type de document :
Article dans une revue
Liste complète des métadonnées

Littérature citée [42 références]  Voir  Masquer  Télécharger

https://hal-enpc.archives-ouvertes.fr/hal-01815826
Contributeur : Romain Boistel <>
Soumis le : mercredi 4 juillet 2018 - 16:34:45
Dernière modification le : vendredi 17 juillet 2020 - 17:08:47
Archivage à long terme le : : lundi 1 octobre 2018 - 13:30:11

Fichier

IJCV-2018-Loing-et-al.pdf
Fichiers produits par l'(les) auteur(s)

Identifiants

Citation

Vianney Loing, Renaud Marlet, Mathieu Aubry. Virtual Training for a Real Application: Accurate Object-Robot Relative Localization without Calibration. International Journal of Computer Vision, Springer Verlag, In press, ⟨10.1007/s11263-018-1102-6⟩. ⟨hal-01815826v2⟩

Partager

Métriques

Consultations de la notice

496

Téléchargements de fichiers

433