Academic Bibliography

 Search 200 years of publications by Ghent University researchers.
Advanced search
1 file | 4.53 MB
Add to list
  1. Search results
  2. Learning to grasp arbitrary household...

Learning to grasp arbitrary household objects from a single demonstration

Elias De Coninck, Tim Verbelen (UGent) , Pieter Van Molle (UGent) , Pieter Simoens (UGent) and Bart Dhoedt (UGent)
(2019) 2019 IEEE/RSJ INTERNATIONAL CONFERENCE ON INTELLIGENT ROBOTS AND SYSTEMS (IROS). In IEEE International Conference on Intelligent Robots and Systems p.2372-2377
Author
Organization
Abstract
Upon the advent of Industry 4.0, collaborative robotics and intelligent automation gain more and more traction for enterprises to improve their production processes. In order to adapt to this trend, new programming, learning and collaborative techniques are investigated. Program-by-demonstration is one of the techniques that aim to reduce the burden of manually programming collaborative robots. However, this is often limited to teaching to grasp at a certain position, rather than grasping a certain object. In this paper, we propose a method that learns to grasp an arbitrary object from visual input. While other learning-based approaches for robotic grasping require collecting a large dataset, manually or automatically labeled in a real or simulated world, our approach requires a single demonstration. We present results on grasping various objects with the Franka Panda collaborative robot after capturing a single image from a wrist mounted RGB camera. From this image we learn a robot controller with a convolutional neural network to adapt to changes in the object's position and rotation with less than 5 minutes of training time on a NVIDIA Titan X GPU, achieving over 90% grasp success rate.

Downloads

  • Download
    (...).pdf
    • full text (Published version)
    • |
    • UGent only
    • |
    • PDF
    • |
    • 4.53 MB

Citation

Please use this url to cite or link to this publication:

MLA
De Coninck, Elias, et al. “Learning to Grasp Arbitrary Household Objects from a Single Demonstration.” 2019 IEEE/RSJ INTERNATIONAL CONFERENCE ON INTELLIGENT ROBOTS AND SYSTEMS (IROS), IEEE, 2019, pp. 2372–77, doi:10.1109/IROS40897.2019.8967638.
APA
De Coninck, E., Verbelen, T., Van Molle, P., Simoens, P., & Dhoedt, B. (2019). Learning to grasp arbitrary household objects from a single demonstration. 2019 IEEE/RSJ INTERNATIONAL CONFERENCE ON INTELLIGENT ROBOTS AND SYSTEMS (IROS), 2372–2377. https://doi.org/10.1109/IROS40897.2019.8967638
Chicago author-date
De Coninck, Elias, Tim Verbelen, Pieter Van Molle, Pieter Simoens, and Bart Dhoedt. 2019. “Learning to Grasp Arbitrary Household Objects from a Single Demonstration.” In 2019 IEEE/RSJ INTERNATIONAL CONFERENCE ON INTELLIGENT ROBOTS AND SYSTEMS (IROS), 2372–77. IEEE. https://doi.org/10.1109/IROS40897.2019.8967638.
Chicago author-date (all authors)
De Coninck, Elias, Tim Verbelen, Pieter Van Molle, Pieter Simoens, and Bart Dhoedt. 2019. “Learning to Grasp Arbitrary Household Objects from a Single Demonstration.” In 2019 IEEE/RSJ INTERNATIONAL CONFERENCE ON INTELLIGENT ROBOTS AND SYSTEMS (IROS), 2372–2377. IEEE. doi:10.1109/IROS40897.2019.8967638.
Vancouver
1.
De Coninck E, Verbelen T, Van Molle P, Simoens P, Dhoedt B. Learning to grasp arbitrary household objects from a single demonstration. In: 2019 IEEE/RSJ INTERNATIONAL CONFERENCE ON INTELLIGENT ROBOTS AND SYSTEMS (IROS). IEEE; 2019. p. 2372–7.
IEEE
[1]
E. De Coninck, T. Verbelen, P. Van Molle, P. Simoens, and B. Dhoedt, “Learning to grasp arbitrary household objects from a single demonstration,” in 2019 IEEE/RSJ INTERNATIONAL CONFERENCE ON INTELLIGENT ROBOTS AND SYSTEMS (IROS), Macau, PEOPLES R CHINA, 2019, pp. 2372–2377.
@inproceedings{8672126,
  abstract     = {{Upon the advent of Industry 4.0, collaborative robotics and intelligent automation gain more and more traction for enterprises to improve their production processes. In order to adapt to this trend, new programming, learning and collaborative techniques are investigated. Program-by-demonstration is one of the techniques that aim to reduce the burden of manually programming collaborative robots. However, this is often limited to teaching to grasp at a certain position, rather than grasping a certain object. In this paper, we propose a method that learns to grasp an arbitrary object from visual input. While other learning-based approaches for robotic grasping require collecting a large dataset, manually or automatically labeled in a real or simulated world, our approach requires a single demonstration. We present results on grasping various objects with the Franka Panda collaborative robot after capturing a single image from a wrist mounted RGB camera. From this image we learn a robot controller with a convolutional neural network to adapt to changes in the object's position and rotation with less than 5 minutes of training time on a NVIDIA Titan X GPU, achieving over 90% grasp success rate.}},
  author       = {{De Coninck, Elias and Verbelen, Tim and Van Molle, Pieter and Simoens, Pieter and Dhoedt, Bart}},
  booktitle    = {{2019 IEEE/RSJ INTERNATIONAL CONFERENCE ON INTELLIGENT ROBOTS AND SYSTEMS (IROS)}},
  isbn         = {{9781728140049}},
  issn         = {{2153-0858}},
  language     = {{eng}},
  location     = {{Macau, PEOPLES R CHINA}},
  pages        = {{2372--2377}},
  publisher    = {{IEEE}},
  title        = {{Learning to grasp arbitrary household objects from a single demonstration}},
  url          = {{http://doi.org/10.1109/IROS40897.2019.8967638}},
  year         = {{2019}},
}
Altmetric
View in Altmetric
Web of Science
Times cited: