The Effect of Scale Consistency between Real and Virtual Spaces on Immersion in Exhibition Hybrid Spaces

2025-07-22

AI Summary

This study addresses the diminished immersion caused by scale inconsistency in exhibition-oriented mixed-reality (MR) environments. It presents the first systematic investigation into how mapping ratios—specifically, the scaling of virtual environments and avatars relative to physical space—affect user immersion. An immersive interactive system was developed using Intel 3D Athlete Tracking for motion capture, integrated with a VR rendering engine and subjective evaluation via Likert-scale questionnaires. Statistical analysis employed ANOVA with Tukey HSD post-hoc tests. Results indicate that a 130% enlarged virtual environment yields the highest user preference; avatar height at 75%–100% of the user’s actual height maximizes immersion, significantly outperforming both undersized and oversized ratios. These findings reveal differential perceptual sensitivity across scale ranges, providing empirically grounded, quantitative guidelines for spatial scaling in MR exhibition design. The framework has been deployed in Intel’s Global Trade Center exhibition hall.

Technology Category

• Humans and AI: Game Design — Virtual Humans, NPCs and Autonomous Characters
• Machine Learning: Scalability of ML Systems
• Cognitive Modeling & Cognitive Systems: Social Cognition And Interaction

Application Category

• User Modeling, Personalization and Recommendation: Practical large-scale studies of user experience
• Security and Privacy: Large-scale security measurements
• Search and Retrieval-Augmented AI: Web evaluation methodologies and metrics

Abstract

In exhibition hybrid spaces, scale consistency between real and virtual spaces is crucial for user immersion. However, there is currently a lack of systematic research to determine appropriate virtual-to-real mapping ratios. This study developed an immersive interaction system based on Intel 3D Athlete Tracking body mapping technology. Two experiments investigated the impact of virtual space and virtual avatar scale on immersion. Experiment 1 investigated 30 participants' preferences for virtual space scale, while Experiment 2 tested the effect of 6 different virtual avatar sizes (25%-150%) on immersion. A 5-point Likert scale was used to assess immersion, followed by analysis of variance and Tukey HSD post-hoc tests. Experiment 1 showed that participants preferred a virtual space ratio of 130% (mean 127.29%, SD 8.55%). Experiment 2 found that virtual avatar sizes within the 75%-100% range produced optimal immersion (p < 0.05). Immersion decreased significantly when virtual avatar sizes deviated from users' actual height (below 50% or above 125%). Participants were more sensitive to size changes in the 25%-75% range, while perception was weaker for changes in the 75%-100% range. Virtual environments slightly larger than real space (130%) and virtual avatars slightly smaller than users (75%-100%) optimize user immersion. These findings have been applied in the Intel Global Trade Center exhibition hall, demonstrating actionable insights for designing hybrid spaces that enhance immersion and coherence.

Problem

• Investigates scale consistency impact on immersion in hybrid spaces
• Determines optimal virtual-to-real mapping ratios for user immersion
• Tests virtual avatar size effects on immersion perception

Innovation

• Intel 3D Athlete Tracking body mapping
• 130% virtual space ratio preferred
• 75%-100% virtual avatar size optimal