Abstract
Recent advances in neural rendering have created new possibilities for constructing highly realistic and efficient virtual environments. Among them, 3D Gaussian Splatting (3DGS) has emerged as a promising technique, combining photorealistic reconstruction with real-time rendering performance. This study evaluates the potential of 3DGS in virtual reality (VR) systems by comparing it against traditional 3D modeling and 360-degree panoramic imaging. We adopt a dual-pathway user experience framework, in which high-fidelity environments influence behavioral outcomes through two distinct routes: an affective route, grounded in immersion and emotional resonance, and a cognitive route, based on information credibility and evaluative judgment. Experimental results show that 3DGS-based VR most strongly activates the affective pathway, enhancing immersion and emotional engagement, while the cognitive pathway provides a reinforcing influence. These findings demonstrate how rendering fidelity translates into measurable differences in user perception and behavioral intention. Beyond user studies, this research highlights the potential of 3DGS for scalable, high-performance VR applications, providing guidance for both system developers and applied domains such as digital heritage and cultural tourism.
Funding Information
  • 2024 Ministry of Education of China Humanities and Social Sciences Research Planning Fund Project (24YJAZH139)

This publication has 39 references indexed in Scilit: