Disentangled Relational Representations for Explaining and Learning from Demonstration

Yordan Hristov, Daniel Angelov, Michael Burke, Alex Lascarides, Subramanian Ramamoorthy

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

Learning from demonstration is an effective method for human users to instruct desired robot behaviour. However, for most non-trivial tasks of practical interest, efficient learning from demonstration depends crucially on inductive bias in the chosen structure for rewards/costs and policies. We address the case where this inductive bias comes from an exchange with a human user. We propose a method in which a learning agent utilizes the information bottleneck layer of a high-parameter variational neural model, with auxiliary loss terms, in order to ground abstract concepts such as spatial relations. The concepts are referred to in natural language instructions and are manifested in the high-dimensional sensory input stream the agent receives from the world. We evaluate the properties of the latent space of the learned model in a photorealistic synthetic environment and particularly focus on examining its usability for downstream tasks. Additionally, through a series of controlled table-top manipulation experiments, we demonstrate that the learned manifold can be used to ground demonstrations as symbolic plans, which can then be executed on a PR2 robot.
Original languageEnglish
Title of host publicationProceedings of the Conference on Robot Learning 2019
PublisherPMLR
Pages870-884
Number of pages15
Publication statusAccepted/In press - 7 Sep 2019
EventConference on Robot Learning (CoRL) - 2019 Edition - Osaka, Japan
Duration: 30 Oct 20191 Nov 2019
https://www.robot-learning.org/

Publication series

NameProceedings of Machine Learning Research
Volume100
ISSN (Electronic)2640-3498

Conference

ConferenceConference on Robot Learning (CoRL) - 2019 Edition
Abbreviated titleCoRL 2019
Country/TerritoryJapan
CityOsaka
Period30/10/191/11/19
Internet address

Keywords

  • Human-robot interaction
  • Interpretable symbol grounding
  • Learning from demonstration

Fingerprint

Dive into the research topics of 'Disentangled Relational Representations for Explaining and Learning from Demonstration'. Together they form a unique fingerprint.

Cite this