Manuscript received September 10, 2025; revised November 18, 2025; accepted December 23, 2025; published May 27, 2026.

Abstract—To achieve photorealistic Virtual Reality (VR) tours, a balance is required between visual fidelity, responsiveness, and the strict runtime and memory constraints of head-mounted displays. As a more recent explicit view synthesis method, three-Dimensional Gaussian Splatting (3DGS) has been proposed as a real-time alternative to Neural Radiance Fields (NeRF). In this study, we present an overview of 3DGS, demonstrating its application to immersive VR tours, and propose a design philosophy for its implementation on standalone headsets. Specifically, we review state-of-the-art 3DGS variants with respect to photorealism, scalability, efficiency, dynamics, and semantics; analyze VR-specific constraints related to latency, stereo stability, and thermal, memory, and interaction requirements; collate VR style datasets which are suitable for training and benchmarking 3DGS-based pipelines; present case studies in domains such as museums, real estate walkthroughs, and large-scale campus tours; synthesize an evaluation framework that organizes metrics for visual quality, performance, memory usage, and interaction responsiveness. The evaluation framework has four axes: visual fidelity, runtime performance, memory footprint, and interaction responsiveness. Although more recent accelerated NeRF-based systems (e.g., Instant-Neural Graphics Primitives (Instant-NGP) and baked radiance fields) directly reduce the performance difference between 3DGS and explicit methods, 3DGS still provides better trade-offs in interactive tours, such as predictable latency and instant editability. Finally, open challenges in streaming, reflectance, semantics, and authoring ergonomics are outlined, and future directions for 3DGS-based VR tour systems are discussed.

Keywords—gaussian splatting, virtual reality, photorealistic rendering, immersive tours, real-time graphics, evaluative framework

Cite: Md Mridul Hossain and SeongKi Kim, "Gaussian Splatting for Photorealistic Immersive Virtual Tours," Journal of Image and Graphics, Vol. 14, No. 3, pp. 387-402, 2026.

Copyright © 2026 by the authors. This is an open access article distributed under the Creative Commons Attribution License which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited (CC BY 4.0).