GLINT: Modeling Scene-Scale Transparency via Gaussian Radiance Transport

2026-03-27

AI Summary

Existing 3D Gaussian splatting methods struggle to model transparent objects due to their inability to disentangle the radiance contributions of transparent interfaces from the transmitted geometry behind them. This work proposes an explicit Gaussian decomposition representation that separates the primary transparent surface from its reflected and transmitted radiance components. By integrating geometric separation cues with priors from a pre-trained video relighting model, our approach enables end-to-end localization and reconstruction of transparency. Within the 3D Gaussian splatting framework, we introduce radiance decoupling modeling, geometry–material prior guidance, and a bootstrapped optimization strategy for transparency estimation. The method significantly outperforms existing techniques on complex transparent scenes, achieving more accurate and visually consistent reconstructions of both geometry and appearance.

Abstract

While 3D Gaussian splatting has emerged as a powerful paradigm, it fundamentally fails to model transparency such as glass panels. The core challenge lies in decoupling the intertwined radiance contributions from transparent interfaces and the transmitted geometry observed through the glass. We present GLINT, a framework that models scene-scale transparency through explicit decomposed Gaussian representation. GLINT reconstructs the primary interface and models reflected and transmitted radiance separately, enabling consistent radiance transport. During optimization, GLINT bootstraps transparency localization from geometry-separation cues induced by the decomposition, together with geometry and material priors from a pre-trained video relighting model. Extensive experiments demonstrate consistent improvements over prior methods for reconstructing complex transparent scenes.

Problem

• transparency modeling
• radiance transport
• 3D reconstruction
• Gaussian splatting
• scene-scale transparency

Innovation

• Gaussian splatting
• scene-scale transparency
• radiance transport
• decomposed representation
• transparent surface reconstruction