Avatar-Centered Feedback: Dynamic Avatar Alterations Can Induce Avoidance Behaviors to Virtual Dangers
Abstract One singularity of VR is its capacity to generate a strong sensation of being present in a dangerous environment, without risking the physical consequences. However, this absence of consequences when interacting with virtual dangers might also limit the induction of realistic responses, such as avoidance behaviors, which are a key factor of various VR applications (e.g., training, journalism, or exposure therapies). To address this limitation, we propose avatar-centered feedback, a novel, device-free approach, consisting of dynamically altering the avatar’s appearance and movements to offer coherent feedback from interactions with the virtual environment, including dangers. To begin with, we present a design space clarifying the range of potential implementations for this approach. Then, we tested this approach in a metallurgy scenario, where participants’ virtual hands would redden and display burns as they got closer to a virtual fire (appearance alteration), and simulate short withdrawal reflexes movements when a spark burst next to them (movement alteration). Our results show that in comparison to a control group, participants receiving avatar-centered feedback demonstrated significantly more avoidance behaviors to the virtual fire. Interestingly, we also found that experiencing avatar-centered feedback of fire significantly increased avoidance behaviors toward a following danger of different nature (a circular saw). These results suggest that avatar-centered feedback can also impact the general perception of the avatar vulnerability to the virtual environment.
Authors
- Rebecca Fribourg (ORCID: https://orcid.org/0000-0002-4572-477X)
- Ferran Argelaguet (ORCID: https://orcid.org/0000-0002-6160-8015)
- Antonin Cheymol
- Jean‐Marie Normand (ORCID: https://orcid.org/0000-0003-0557-4356)
- Anatole Lécuyer (ORCID: https://orcid.org/0000-0002-1409-244X)
Institutions
- École Centrale de Nantes (FR)
- Centre National de la Recherche Scientifique (FR)
- Institut national de recherche en sciences et technologies du numérique (FR)
- Institut de Recherche en Informatique et Systèmes Aléatoires (FR)
- Laboratoire des Sciences du Numérique de Nantes (FR)
- Université de Rennes (FR)
- Nantes Université (FR)
Publication Details
- Published
- 2024-10-21
- DOI
- https://doi.org/10.1109/ismar62088.2024.00023
- Citations
- 5
- Primary Topic
- Human-Automation Interaction and Safety
- Type
- article
- Field-Weighted Citation Impact
- 3.22