An empirical characterization of parsimonious intention inference for cognitive-level imitation learning

Garrett Katz, Di Wei Huang, Rodolphe Gentili, James Reggia

Research output: Chapter in Book/Entry/PoemConference contribution

2 Scopus citations

Abstract

Imitation learning is a promising route to better collaboration between humans and artificial agents. It will be most effective if the agent has some cognitive-level “understanding” of a human demonstrator’s intentions. Inferring intent is an example of abductive reasoning, wherein an agent explains the available evidence based on causal knowledge. Good explanations should satisfy some notion of parsimony (“Occam’s razor”), but the optimal notion of parsimony is often application-specific. We compare several such notions in the context of intention inference, using a robotic imitation learning scenario and the Monroe County Corpus, a standard benchmark in intention inference. Our results suggest that the most popular notions of parsimony in general are not necessarily appropriate for intention inference in particular.

Original languageEnglish (US)
Title of host publication2017 World Congress in Computer Science, Computer Engineering and Applied Computing, CSCE 2017 - Proceedings of the 2017 International Conference on Artificial Intelligence, ICAI 2017
EditorsHamid R. Arabnia, David de la Fuente, Elena B. Kozerenko, Jose A. Olivas, Fernando G. Tinetti
PublisherCSREA Press
ISBN (Electronic)9781601324603
StatePublished - Jan 1 2017
Externally publishedYes
Event2017 International Conference on Artificial Intelligence, ICAI 2017 at 2017 World Congress in Computer Science, Computer Engineering and Applied Computing, CSCE 2017 - Las Vegas, United States
Duration: Jul 17 2017Jul 20 2017

Publication series

Name2017 World Congress in Computer Science, Computer Engineering and Applied Computing, CSCE 2017 - Proceedings of the 2017 International Conference on Artificial Intelligence, ICAI 2017

Conference

Conference2017 International Conference on Artificial Intelligence, ICAI 2017 at 2017 World Congress in Computer Science, Computer Engineering and Applied Computing, CSCE 2017
Country/TerritoryUnited States
CityLas Vegas
Period7/17/177/20/17

Keywords

  • Artificial intelligence
  • Imitation learning
  • Intention inference
  • Parsimonious covering theory

ASJC Scopus subject areas

  • Software
  • Artificial Intelligence

Fingerprint

Dive into the research topics of 'An empirical characterization of parsimonious intention inference for cognitive-level imitation learning'. Together they form a unique fingerprint.

Cite this