Please use this identifier to cite or link to this item: doi:10.22028/D291-39187
Volltext verfügbar? / Dokumentlieferung
Title: Predicting Mid-Air Interaction Movements and Fatigue Using Deep Reinforcement Learning
Author(s): Cheema, Noshaba
Frey-Law, Laura A.
Naderi, Kourosh
Lehtinen, Jaakko
Slusallek, Philipp
Hämäläinen, Perttu
Language: English
Title: CHI'20 : Proceedings of the 2020 CHI Conference on Human Factors in Computing Systems, April 25-30, 2020, Honolulu, HI, USA
Pages: 1-13
Publisher/Platform: ACM
Year of Publication: 2020
Place of publication: New York
Place of the conference: Honolulu, HI, USA
DDC notations: 004 Computer science, internet
Publikation type: Conference Paper
Abstract: A common problem of mid-air interaction is excessive arm fatigue, known as the "Gorilla arm" effect. To predict and prevent such problems at a low cost, we investigate user testing of mid-air interaction without real users, utilizing biomechanically simulated AI agents trained using deep Reinforcement Learning (RL). We implement this in a pointing task and four experimental conditions, demonstrating that the simulated fatigue data matches human fatigue data. We also compare two effort models: 1) instantaneous joint torques commonly used in computer animation and robotics, and 2) the recent Three Compartment Controller (3CC-) model from biomechanical literature. 3CC- yields movements that are both more efficient and relaxed, whereas with instantaneous joint torques, the RL agent can easily generate movements that are quickly tiring or only reach the targets slowly and inaccurately. Our work demonstrates that deep RL combined with the 3CC- provides a viable tool for predicting both interaction movements and user experiencein silico, without users.
DOI of the first publication: 10.1145/3313831.3376701
URL of the first publication:
Link to this record: urn:nbn:de:bsz:291--ds-391879
ISBN: 978-1-4503-6708-0
Date of registration: 1-Mar-2023
Faculty: MI - Fakultät für Mathematik und Informatik
Department: MI - Informatik
Professorship: MI - Prof. Dr. Philipp Slusallek
Collections:SciDok - Der Wissenschaftsserver der Universität des Saarlandes

Files for this record:
There are no files associated with this item.

Items in SciDok are protected by copyright, with all rights reserved, unless otherwise indicated.