analiza aiqbench
Ref-DGS addresses the computational bottleneck in modeling near-field specular reflections by decoupling geometry from reflection within a rasterization-based pipeline. This approach eliminates the need for expensive explicit ray tracing while maintaining state-of-the-art performance on reflective scenes through a dual Gaussian representation and adaptive mixing shader. The method offers a significant efficiency gain for novel view synthesis tasks where existing Gaussian splatting techniques struggle with strong specular highlights.
Ref-DGS: Reflective Dual Gaussian Splatting
arXiv:2603.07664v3 Announce Type: replace-cross Abstract: The reflective appearance, especially strong and typically near-field specular reflections, poses a fundamental challenge for accurate surface reconstruction and novel view synthesis. Existing Gaussian splatting methods either fail to model near-field specular reflections or rely on explicit ray tracing at substantial computational cost. We present \textbf{Ref-DGS}, a reflective dual Gaussian splatting framework that addresses this trade-off by decoupling surface reconstruction from specular reflection within an efficient rasterization-based pipeline. Ref-DGS introduces a dual Gaussian scene representation consisting of geometry Gaussians and complementary local reflection Gaussians that capture near-field specular interactions without explicit ray tracing, along with a global environment reflection field for modeling far-field specular reflections. To predict specular radiance, we further propose a lightweight, physically-aware specular adaptive mixing shader that fuses global and local specular features. Experiments demonstrate that Ref-DGS achieves state-of-the-art performance on reflective scenes while training substantially faster than ray-based Gaussian methods.
Abstract:The reflective appearance, especially strong and typically near-field specular reflections, poses a fundamental challenge for accurate surface reconstruction and novel view synthesis. Existing Gaussian splatting methods either fail to model near-field specular reflections or rely on explicit ray tracing at substantial computational cost. We present \textbf{Ref-DGS}, a reflective dual Gaussian splatting framework that addresses this trade-off by decoupling surface reconstruction from specular reflection within an efficient rasterization-based pipeline. Ref-DGS introduces a dual Gaussian scene representation consisting of geometry Gaussians and complementary local reflection Gaussians that capture near-field specular interactions without explicit ray tracing, along with a global environment reflection field for modeling far-field specular reflections. To predict specular radiance, we further propose a lightweight, physically-aware specular adaptive mixing shader that fuses global and local specular features. Experiments demonstrate that Ref-DGS achieves state-of-the-art performance on reflective scenes while training substantially faster than ray-based Gaussian methods.
Computer Vision and Pattern Recognition (cs.CV); Artificial Intelligence (cs.AI); Graphics (cs.GR)
arXiv:2603.07664 [cs.CV]
(or arXiv:2603.07664v3 [cs.CV] for this version)
Text integral sub licența deschisă a publicației · 163 cuvinte
Subiecte conexe
[AINews] not much happened today
Latent Space · 18 sept. 2026
SpotlessGS: Relightable 3D Gaussian Splatting under Dynamic Illumination for Robotic Perception
arXiv · 18 aug. 2026
Dual Covariance Gaussian Splatting SLAM: Decoupling Rendering and Registration for Robust Real-Time Tracking
arXiv · 24 sept. 2026
CoRef-GS: Cooperative Referring Gaussian Splatting for Multi-Agent Scene Understanding
arXiv · 18 sept. 2026