Observing Virtual Avatars: The Impact of Avatars’ Fidelity on Identifying Interactions
Florian Mathis, Kami E. Vaniea, Mohamed A. Khamis
Abstract
Open-access reader
Florian Mathis, Kami E. Vaniea, Mohamed A. Khamis
Abstract
Open-access reader
There are many cases where observing the interactions of a virtual avatar can be useful. However, it is unclear to what extent the avatar fidelity affects bystanders’ performance when observing gestures performed by avatars. We, therefore, conducted an online study (N=28) with different avatars performing touch, mid-air, and eye gaze gestures. Our avatars range from an abstract avatar to two more realistic ones. Our study shows that an abstract avatar provides bystanders with the same interaction information as more realistic avatars. This implies that it is sufficient to use abstract avatar designs when observing interaction behaviour in virtual environments and researchers do not necessarily need to go through time-consuming and expensive implementations of highly realistic avatars to answer their research questions, making Virtual Reality (VR) studies more accessible to the broader research community. Finally, we discuss studying VR security systems as a potential research application for which abstract avatars can be leveraged.
OpenAlex reports 17 citations for this work. Citation counts describe recorded attention and do not establish research quality.
A contribution statement is not available in the OpenAlex record.
Method details are not available in the OpenAlex metadata.
Findings are not separately available in the OpenAlex metadata.
Limitations are not available in the OpenAlex metadata.
Application details are not available in the OpenAlex metadata.
There are many cases where observing the interactions of a virtual avatar can be useful. However, it is unclear to what extent the avatar fidelity affects bystanders’ performance when observing gestures performed by avatars. We, therefore, conducted an online study (N=28) with different avatars performing touch, mid-air, and eye gaze gestures. Our avatars range from an abstract avatar to two more realistic ones. Our study shows that an abstract avatar provides bystanders with the same interaction information as more realistic avatars. This implies that it is sufficient to use abstract avatar designs when observing interaction behaviour in virtual environments and researchers do not necessarily need to go through time-consuming and expensive implementations of highly realistic avatars to answer their research questions, making Virtual Reality (VR) studies more accessible to the broader research community. Finally, we discuss studying VR security systems as a potential research application for which abstract avatars can be leveraged.
Key concepts: Avatar, Gesture, Human–computer interaction, Computer science, Fidelity, Virtual reality, Implementation, Virtual machine