Walking Animationsystem for Social Extended Reality
This project is already completed.
Background
Motivation and Related Work
Applications for social interaction in virtual worlds are one of the promising application areas of virtual reality (VR). Within social virtual reality applications, users are represented by avatars, that are in most cases cut off below the hip, since plausible leg animations are still very difficult to realize. Some solutions require atleast 3 additional trackers (Wolf et al., 2020) and use complex mathematics such as inverse kinematics (Aristidou et al., 2018) or deep learning approaches (Jiang et al., 2022; Winkler et al., 2022; Ponton et al. 2022).
In distributed social virtual reality applications, the user is unaware of the other users pose, so walking motions are sometimes represented by, classic computer animations. While these computer animations have been widely researched, there is a large gap in research on the effect of walking and gait animations on other users in terms of plausibility and humaness (Thompson et al., 2011).
Goal
The goal of this project is to implement a simple character animation system with motion blending in the Unity Engine using build in functionality such as Animation Systems and Blend Trees with the focus on walking- and gaitsimulation.
This developed animation controller will then be evaluated in a user study with 3 conditions (no gait animation, gait animation without motion blending and gait animation with motion blending) to compare the effect of the different gait animations in terms of plausibility and humanity.
Tasks
The topic will focus on the following tasks:
- Literatur and state-of-the-art research
- Determination of an animation library for gait animation of virtual human agents
- Implementation of a simple animation system with motion blending for gait animations
- Comparative user study
- Evaluation and presentation of results
Prerequisites
- Experience with Unity game engine
- First touch points with 3D animations
- Affinity for technology
Contact Persons at the University Würzburg
Christopher Göttfert (Primary Contact Person)Mensch-Computer-Interaktion, Universität Würzburg
christopher.goettfert@uni-wuerzburg.de
Marc Erich Latoschik
Mensch-Computer-Interaktion, Universität Würzburg
marc.latoschik@uni-wuerzburg.de