Hier finden Sie wissenschaftliche Publikationen aus den Fraunhofer-Instituten.

Model-Free Grasp Learning Framework based on Physical Simulation

: Riedlinger, Marc A.; Völk, Markus; Kleeberger, Kilian; Khalid, Muhammad Usman; Bormann, Richard

Informationstechnische Gesellschaft -ITG-; Verband Deutscher Maschinen- und Anlagenbau e.V. -VDMA-, Frankfurt/Main; Verband der Elektrotechnik, Elektronik, Informationstechnik -VDE-:
ISR 2020, 52nd International Symposium on Robotics : December, 9-10, 2020, Online-Event. In conjunction with Automatica (abgesagt), December 8-11, 2020, Munich, CD-ROM
Berlin: VDE-Verlag, 2020
ISBN: 978-3-8007-5428-1 (Print)
ISBN: 978-3-8007-5429-8 (Online)
ISBN: 3-8007-5428-2
International Symposium on Robotics (ISR) <52, 2020, Online>
Conference Paper
Fraunhofer IPA ()
Robotik; deep grasping; lernendes System; Simulation

The work at hand presents a generic framework to build classifiers that allow to predict the quality of 6-DOF grasp candidates for arbitrary mechanical grippers based on the depth data captured by a depth sensor. Hereby, the framework covers the whole process of setting up a deep neural network for a given mechanical gripper by making use of synthetic data resulting from a new grasp simulation tool. Furthermore, a new extended convolutional neural network (CNN) architecture is introduced that estimates the quality of a suggested grasp candidate based on local depth information and the pose of the corresponding grasp. As a result, robust grasp candidates can be detected in a model-free fashion.