Empirically Evaluating the Effects of Eye Height and Self-Avatars on Dynamic Passability Affordances in Virtual Reality

• Published in: 2023 IEEE Conference Virtual Reality and 3D User Interfaces (VR) pp. 308 - 317
• Main Authors: Bhargava, Ayush, Venkatakrishnan, Roshan, Venkatakrishnan, Rohith, Solini, Hannah, Lucaites, Kathryn, Robb, Andrew C., Pagano, Christopher C., Babu, Sabarish V.
• Format: Conference Proceeding
• Language: English
• Published: IEEE 01-03-2023
• Subjects: Affordances; Avatars; Empirical-studies-in-HCI; External stimuli; Games; Human-centered-computing; Roads; Three-dimensional displays; Virtual environments
• ISSN: 2642-5254
• DOI: 10.1109/VR55154.2023.00046

Over the past two decades self-avatars have been shown to affect the perception of both oneself and of environmental properties including the sizes and distances of elements in immersive virtual environments. However, virtual avatars that accurately match the body proportions of their users remain inaccessible to the general public. As such, most virtual experiences that represent the user have a generic avatar that does not fit the proportions of the users' body. This can negatively affect judgments involving affordances, such as passability and maneuverability, which pertain to the relationship between the properties of environmental elements relative to the properties of the user providing information about actions that can be enacted. This is especially true when the task requires the user to maneuver around moving objects like in games. Therefore, it is necessary to understand how different sized self-avatars affect the perception of affordances in dynamic virtual environments. To better understand this, we conducted an experiment investigating how a self-avatar that is either the same size, 20% shorter, or 20% taller, than the user's own body affects passability judgments in a dynamic virtual environment. Our results suggest that the presence of self-avatars results in better regulatory and safer road crossing behavior, and helps participants synchronize self-motion to external stimuli quicker than in the absence of self-avatars.