|arXiv ID:||2211.10168||Title:||Language-conditioned reinforcement learning to solve misunderstandings with action corrections||Language:||English||Authors:||Röder, Frank
|Keywords:||reinforcement learning; instruction-following; action correction; misunderstanding; ambiguity; negation||Issue Date:||17-Nov-2022||Source:||36th Conference on Neural Information Processing Systems (NeurIPS 2022)||Abstract (english):||
Human-to-human conversation is not just talking and listening. It is an incremental process where participants continually establish a common understanding to rule out misunderstandings. Current language understanding methods for intelligent robots do not consider this. There exist numerous approaches considering non-understandings, but they ignore the incremental process of resolving misunderstandings. In this article, we present a first formalization and experimental validation of incremental action-repair for robotic instruction-following based on reinforcement learning. To evaluate our approach, we propose a collection of benchmark environments for action correction in language-conditioned reinforcement learning, utilizing a synthetic instructor to generate language goals and their corresponding corrections. We show that a reinforcement learning agent can successfully learn to understand incremental corrections of misunderstood instructions.
|Conference:||36th Conference on Neural Information Processing Systems, NeurIPS 2022||URI:||http://hdl.handle.net/11420/14361||Institute:||Data Science Foundations E-21||Document Type:||Conference Paper (not in Proceedings)||Project:||Lernen von konversationaller Aktionsreparatur für intelligente Roboter
SPP 2134: Modellierung des peripersonalen Raumes und Körperschemas eines Roboters für adaptives Lernen und Imitation
SPP 2134: Ideomotor Transfer for Active Self-Emergence
|Funded by:||Deutsche Forschungsgemeinschaft (DFG)||Peer Reviewed:||Yes||Cites:||https://openreview.net/forum?id=lWd0qiv9E-|
|Appears in Collections:||Publications without fulltext|
Show full item record
checked on Mar 10, 2023
Add Files to Item
Note about this record
Items in TORE are protected by copyright, with all rights reserved, unless otherwise indicated.