Publication
Motion Capturing in Large-Scale VR Environments using Consumer Hardware
Janis Sprenger; Lorena Hell; Matthias Klusch
In: Proceedings of the 17th International Symposium of 3-D Analysis of Human Movement. International Symposium of 3-D Analysis of Human Movement (3D-AHM-2022), 17th, July 16-19, Tokyo, Japan, International Society of Biomechanics, 2022.
Abstract
There are many use-cases where motion capturing is required but not physically possible in real life. These usecases
include pedestrian crossings with autonomous vehicles, shop-floor analysis of workers inside a digital
factory prototype, and analysis of behavior in novel architectural environments. In these cases, simulation with a
virtual reality is beneficial. At the same time, an accurate capturing of the motions and the behavior are required
for evaluation and training of new digital agent models or ergonomic assessment. Pedestrians, for example, should
only cross a limited number of streets to provide enough data to train digital agent models, which in turn are
utilized to provide accurate and natural synthetic training environments for autonomous vehicles. Factory workers,
on the other hand, should interact with a new shop-floor design, and their ergonomics should be evaluated to
optimize the design before building the actual station.
In this work, we present our setup, processing, and experience of conducting motion capturing in large-scale VR
setups. Test subjects were recorded in a virtual scene of an urban street while walking along the street and crossing
the street. Although the resulting data is of very high quality, we can report several challenges and implications
for future studies with the aim to assist others in the planning and implementation of similar motion capture and
VR setups.
Projekte
MMISS - Multimedia - Instruktion in Sicheren Systemen,
REACT - Autonomous Driving: Modeling, learning and simulation environment for pedestrian behavior in critical traffic situations
REACT - Autonomous Driving: Modeling, learning and simulation environment for pedestrian behavior in critical traffic situations