Robotic grasp pose detection using deep learning

Authors

Shehan Caldera, Edith Cowan UniversityFollow
Alexander Rassau, Edith Cowan UniversityFollow
Douglas Chai, Edith Cowan UniversityFollow

Document Type

Conference Proceeding

Publisher

Institute of Electrical and Electronics Engineers Inc.

School

School of Engineering

RAS ID

28394

Comments

Caldera, S., Rassau, A., & Chai, D. (2018, November). Robotic Grasp Pose Detection Using Deep Learning. In 2018 15th International Conference on Control, Automation, Robotics and Vision (ICARCV) (pp. 1966-1972). IEEE. Available here

Abstract

Recent advancements in Deep Learning have accelerated the capabilities of robotic systems in terms of visual perception, object manipulation, automated navigation, and human-robot collaboration. This paper proposes the use of a transfer learning technique with deep convolutional neural networks to learn how to visually identify the grasping configurations for a parallel plate gripper that will be used to grasp various household objects. The Red-Green-Blue-Depth (RGB-D) data from the Cornell Grasp Dataset is used to train the network model using an end-to-end learning method. With this method, we achieve a grasping configuration prediction accuracy of 93.91%.

DOI

10.1109/ICARCV.2018.8581091

Related Publications

Caldera, S. (2019). Learning to grasp in unstructured environments with deep convolutional neural networks using a Baxter Research Robot. Retrieved from https://ro.ecu.edu.au/theses/2170

Access Rights

subscription content