发表机构
ZJU-UIUC Institute, Zhejiang University; College of Mathematics, Sichuan University; College of Computer Science and Technology, Zhejiang University; Chengdu Minto Tech（浙江大学ZJU-UIUC学院; 四川大学数学学院; 浙江大学计算机科学与技术学院; 成都敏途科技）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
LightFuse是基于2D高斯框架的交互式场景重建方法，通过多扫描融合、几何细化与材质-光照分解，实现可重光照的可编辑场景，在合成场景上重光照质量优于现有基线。
AI 中文摘要
可重光照交互式场景重建旨在从不同物体布置的扫描数据构建可编辑的3D模型，并在新光照下渲染新布局。现有方法要么将光照烘焙到外观中，要么仅为固定场景恢复材质与光照，导致编辑后的布局出现阴影和间接光照不一致的问题。本文提出LightFuse，这是一个2D高斯框架，通过显式材质-光照分解和基于物理的重光照扩展了交互式场景重建能力。LightFuse首先融合不同状态下的观测数据以重建共享背景与可移动物体；接着进行面向光线追踪的几何细化，生成更完整、一致的表面；在细化后的几何上，采用带可微单次反弹光线追踪的分阶段训练，将共享的金属-粗糙度材质与特定状态的环境光照分离开。生成的场景支持物体重新布置、材质编辑和重光照，每次交互后通过光线追踪重新计算外观。在合成场景上的实验表明，其重光照质量达到当前最优水平，平均比最强基线方法高出9.74 dB的PSNR和0.121的SSIM。项目页面：this https URL
英文摘要
Relightable interactive scene reconstruction aims to build an editable 3D model from scans of different object arrangements and render new layouts under novel illumination. Existing methods either bake lighting into appearance or recover material and illumination only for fixed scenes, leaving edited layouts with inconsistent shadows and indirect lighting. We present LightFuse, a 2D Gaussian framework that extends interactive scene reconstruction with explicit material-illumination decomposition and physically based relighting. LightFuse first fuses observations across states to reconstruct a shared background and movable objects. It then conducts ray-tracing-oriented geometry refinement to produce more complete and consistent surfaces. On the refined geometry, staged training with differentiable one-bounce ray tracing separates shared metallic--roughness material from state-specific environment lighting. The resulting scene supports object rearrangement, material editing, and relighting, while ray tracing recomputes appearance after each interaction. Experiments across synthetic scenes demonstrate state-of-the-art relighting quality, outperforming the strongest baseline by +9.74\,dB PSNR and +0.121 SSIM on average. Project page: https://zhn202.github.io/LightFuse/
