We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:

Download:

Current browse context:

cs.RO

Change to browse by:

cs

References & Citations

DBLP - CS Bibliography

Bookmark

(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo ScienceWISE logo

Computer Science > Robotics

Title: Disturbance-Injected Robust Imitation Learning with Task Achievement

Abstract: Robust imitation learning using disturbance injections overcomes issues of limited variation in demonstrations. However, these methods assume demonstrations are optimal, and that policy stabilization can be learned via simple augmentations. In real-world scenarios, demonstrations are often of diverse-quality, and disturbance injection instead learns sub-optimal policies that fail to replicate desired behavior. To address this issue, this paper proposes a novel imitation learning framework that combines both policy robustification and optimal demonstration learning. Specifically, this combinatorial approach forces policy learning and disturbance injection optimization to focus on mainly learning from high task achievement demonstrations, while utilizing low achievement ones to decrease the number of samples needed. The effectiveness of the proposed method is verified through experiments using an excavation task in both simulations and a real robot, resulting in high-achieving policies that are more stable and robust to diverse-quality demonstrations. In addition, this method utilizes all of the weighted sub-optimal demonstrations without eliminating them, resulting in practical data efficiency benefits.
Comments: 7 pages, Accepted by the 2022 International Conference on Robotics and Automation (ICRA 2022)
Subjects: Robotics (cs.RO)
Cite as: arXiv:2205.04195 [cs.RO]
  (or arXiv:2205.04195v1 [cs.RO] for this version)

Submission history

From: Hirotaka Tahara [view email]
[v1] Mon, 9 May 2022 11:25:12 GMT (12741kb,D)

Link back to: arXiv, form interface, contact.