TY - GEN
T1 - Modelling competitive human action using dynamical motor primitives for the development of human-like artificial agents
AU - Ekdawi, Sarah
AU - Patil, Gaurav
AU - Kallen, Rachel W.
AU - Richardson, Michael J.
N1 - Copyright the Author(s) 2022. Version archived for private and non-commercial use with the permission of the author/s and according to publisher conditions. For further rights please contact the publisher.
PY - 2022
Y1 - 2022
N2 - With artificial intelligence technologies becoming commonplace today, enhancing the efficiency of human-artificial agent (AA) interactions has become increasingly important. A growing body of research has revealed how dynamic motor primitives (DMPs) of human perceptual-motor behavior can be used to create ‘human-like’ AAs, primarily focusing on cooperative tasks. Using air hockey as a representative task, the current experiment is the first part of a large study aimed at determining the utility of DMP-based models for developing ‘human-like’ competitive AAs. Participants played against a preliminary DMP model and the differences in their behaviors were analyzed. Based on these observed differences, a revised model is proposed, with preliminary results revealing that the new model exhibits behaviors more consistent with those of humans. A major implication of this work is that it presents a framework for creating ‘human-like’ AAs that capture the essential human decision and movement dynamics without requiring large human gameplay datasets.
AB - With artificial intelligence technologies becoming commonplace today, enhancing the efficiency of human-artificial agent (AA) interactions has become increasingly important. A growing body of research has revealed how dynamic motor primitives (DMPs) of human perceptual-motor behavior can be used to create ‘human-like’ AAs, primarily focusing on cooperative tasks. Using air hockey as a representative task, the current experiment is the first part of a large study aimed at determining the utility of DMP-based models for developing ‘human-like’ competitive AAs. Participants played against a preliminary DMP model and the differences in their behaviors were analyzed. Based on these observed differences, a revised model is proposed, with preliminary results revealing that the new model exhibits behaviors more consistent with those of humans. A major implication of this work is that it presents a framework for creating ‘human-like’ AAs that capture the essential human decision and movement dynamics without requiring large human gameplay datasets.
KW - dynamic motor primitives
KW - human-machine interaction
KW - human behavior modelling
KW - artificial intelligence
UR - http://www.scopus.com/inward/record.url?scp=85146434589&partnerID=8YFLogxK
M3 - Conference proceeding contribution
T3 - Proceedings of the Annual Meeting of the Cognitive Science Society
SP - 2731
EP - 2737
BT - CogSci2022
A2 - Culbertson, Jennifer
A2 - Perfors, Andrew
A2 - Rabagliati, Hugh
A2 - Ramenzoni, Veronica
PB - Cognitive Science Society
CY - Austin, Texas
T2 - Annual Meeting of the Cognitive Science Society (44th : 2022)
Y2 - 27 July 2022 through 30 July 2022
ER -