TU Darmstadt / ULB / TUprints

Deep Adversarial Reinforcement Learning for Object Disentangling

Laux, Melvin ; Arenz, Oleg ; Peters, Jan ; Pajarinen, Joni (2022):
Deep Adversarial Reinforcement Learning for Object Disentangling. (Postprint)
In: 2020 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), pp. 5504-5510,
Darmstadt, IEEE, 2020 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), Las Vegas, USA (Virtual), 25.10.-29.10.2020, e-ISSN 2153-0866, ISBN 978-1-7281-6212-6,
DOI: 10.26083/tuprints-00022926,
[Conference or Workshop Item]

[img] Text
deep_adversarial_reinforcement_learning_for_object_disentangling.pdf
Copyright Information: In Copyright.

Download (2MB)
Item Type: Conference or Workshop Item
Origin: Secondary publication service
Status: Postprint
Title: Deep Adversarial Reinforcement Learning for Object Disentangling
Language: English
Abstract:

Deep learning in combination with improved training techniques and high computational power has led to recent advances in the field of reinforcement learning (RL) and to successful robotic RL applications such as in-hand manipulation. However, most robotic RL relies on a well known initial state distribution. In real-world tasks, this information is however often not available. For example, when disentangling waste objects the actual position of the robot w.r.t. the objects may not match the positions the RL policy was trained for. To solve this problem, we present a novel adversarial reinforcement learning (ARL) framework. The ARL framework utilizes an adversary, which is trained to steer the original agent, the protagonist, to challenging states. We train the protagonist and the adversary jointly to allow them to adapt to the changing policy of their opponent. We show that our method can generalize from training to test scenarios by training an end-to-end system for robot control to solve a challenging object disentangling task. Experiments with a KUKA LBR+ 7-DOF robot arm show that our approach outperforms the baseline method in disentangling when starting from different initial states than provided during training.

Book Title: 2020 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS)
Place of Publication: Darmstadt
Publisher: IEEE
Collation: 7 ungezählte Seiten
Uncontrolled Keywords: Training, Visualization, Sensitivity, Shape, Reinforcement learning, Task analysis, Robots
Classification DDC: 000 Allgemeines, Informatik, Informationswissenschaft > 004 Informatik
Divisions: 20 Department of Computer Science > Intelligent Autonomous Systems
Event Title: 2020 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS)
Event Location: Las Vegas, USA (Virtual)
Event Dates: 25.10.-29.10.2020
Date Deposited: 25 Nov 2022 12:45
Last Modified: 11 Jan 2023 14:32
DOI: 10.26083/tuprints-00022926
Corresponding Links:
URN: urn:nbn:de:tuda-tuprints-229264
URI: https://tuprints.ulb.tu-darmstadt.de/id/eprint/22926
PPN: 503350842
Export:
Actions (login required)
View Item View Item