PrismGS: Physically-Grounded Anti-Aliasing for High-Fidelity Large-Scale 3D Gaussian Splatting

2025-10-09

AI Summary

In large-scale urban scenes, 3D Gaussian Splatting (3DGS) suffers from severe aliasing, flickering, and optimization instability under high-resolution (e.g., 4K) rendering—primarily due to a mismatch between the fixed Gaussian primitive scale and the inherent multi-scale geometry of urban environments. Method: We propose the first physics-aware regularization framework for 3DGS, introducing pyramid-based pre-filtering supervision, multi-scale consistency loss, and a physically grounded minimum-scale constraint to enforce intrinsic anti-aliasing at the representation level—without modifying the core rendering pipeline or disrupting existing 3DGS workflows. Contribution/Results: Our method significantly improves rendering fidelity and geometric stability on MatrixCity, Mill-19, and UrbanScene3D. It achieves ≈1.5 dB PSNR gain over CityGaussian and maintains high-quality, robust 4K rendering—demonstrating superior scalability and generalization for complex urban reconstruction.

Technology Category

• Computer Vision: 3D Computer Vision
• Search and Optimization: Sampling/Simulation-based Search
• Planning, Routing, and Scheduling: Replanning and Plan Repair

Application Category

• Search and Retrieval-Augmented AI: Efficiency and scalability of Web search engines
• Systems and Infrastructure for Web, Mobile and WoT: Sustainability and carbon-aware systems for Web, mobile, and WoT
• Security and Privacy: Large-scale security measurements

Abstract

3D Gaussian Splatting (3DGS) has recently enabled real-time photorealistic rendering in compact scenes, but scaling to large urban environments introduces severe aliasing artifacts and optimization instability, especially under high-resolution (e.g., 4K) rendering. These artifacts, manifesting as flickering textures and jagged edges, arise from the mismatch between Gaussian primitives and the multi-scale nature of urban geometry. While existing ``divide-and-conquer'' pipelines address scalability, they fail to resolve this fidelity gap. In this paper, we propose PrismGS, a physically-grounded regularization framework that improves the intrinsic rendering behavior of 3D Gaussians. PrismGS integrates two synergistic regularizers. The first is pyramidal multi-scale supervision, which enforces consistency by supervising the rendering against a pre-filtered image pyramid. This compels the model to learn an inherently anti-aliased representation that remains coherent across different viewing scales, directly mitigating flickering textures. This is complemented by an explicit size regularization that imposes a physically-grounded lower bound on the dimensions of the 3D Gaussians. This prevents the formation of degenerate, view-dependent primitives, leading to more stable and plausible geometric surfaces and reducing jagged edges. Our method is plug-and-play and compatible with existing pipelines. Extensive experiments on MatrixCity, Mill-19, and UrbanScene3D demonstrate that PrismGS achieves state-of-the-art performance, yielding significant PSNR gains around 1.5 dB against CityGaussian, while maintaining its superior quality and robustness under demanding 4K rendering.

Problem

• Addressing aliasing artifacts in large-scale 3D Gaussian Splatting
• Resolving flickering textures and jagged edges in urban rendering
• Improving optimization stability for high-resolution 3D scene reconstruction

Innovation

• Pyramidal multi-scale supervision for anti-aliased representation
• Explicit size regularization with physical lower bounds
• Plug-and-play framework compatible with existing pipelines