Skip to main content

Research Repository

Advanced Search

Confidence in uncertainty: Error cost and commitment in early speech hypotheses

Loth, Sebastian; Jettka, Katharina; Giuliani, Manuel; Kopp, Stefan; de Ruiter, Jan

Confidence in uncertainty: Error cost and commitment in early speech hypotheses Thumbnail


Authors

Sebastian Loth

Katharina Jettka

Manuel Giuliani Manuel.Giuliani@uwe.ac.uk
Co- Director Bristol Robotics Laboratory

Stefan Kopp

Jan de Ruiter



Abstract

© 2018 Loth et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited. Interactions with artificial agents often lack immediacy because agents respond slower than their users expect. Automatic speech recognisers introduce this delay by analysing a user’s utterance only after it has been completed. Early, uncertain hypotheses of incremental speech recognisers can enable artificial agents to respond more timely. However, these hypotheses may change significantly with each update. Therefore, an already initiated action may turn into an error and invoke error cost. We investigated whether humans would use uncertain hypotheses for planning ahead and/or initiating their response. We designed a Ghost-in-the-Machine study in a bar scenario. A human participant controlled a bartending robot and perceived the scene only through its recognisers. The results showed that participants used uncertain hypotheses for selecting the best matching action. This is comparable to computing the utility of dialogue moves. Participants evaluated the available evidence and the error cost of their actions prior to initiating them. If the error cost was low, the participants initiated their response with only suggestive evidence. Otherwise, they waited for additional, more confident hypotheses if they still had time to do so. If there was time pressure but only little evidence, participants grounded their understanding with echo questions. These findings contribute to a psychologically plausible policy for human-robot interaction that enables artificial agents to respond more timely and socially appropriately under uncertainty.

Journal Article Type Article
Acceptance Date Jul 17, 2018
Online Publication Date Aug 1, 2018
Publication Date Aug 1, 2018
Deposit Date Aug 6, 2018
Publicly Available Date Aug 6, 2018
Journal PLoS ONE
Electronic ISSN 1932-6203
Publisher Public Library of Science
Peer Reviewed Peer Reviewed
Volume 13
Issue 8
Article Number e0201516
DOI https://doi.org/10.1371/journal.pone.0201516
Keywords robotics, human-robot interaction, human factors, speech recognition
Public URL https://uwe-repository.worktribe.com/output/863319
Publisher URL http://dx.doi.org/10.1371/journal.pone.0201516
Related Public URLs https://doi.org/10.1371/journal.pone.0201516
Contract Date Aug 6, 2018

Files






You might also like



Downloadable Citations