PhysIR-Splat：3D 高斯泼溅中物理一致的热红外辐射传输
PhysIR-Splat: Physically Consistent Thermal Infrared Radiative Transfer in 3D Gaussian Splatting
把温度、发射率与环境辐照度显式绑定到高斯原语上，并在渲染时联合建模热辐射、反射分量与大气透过率，为热红外 NVS 提供可复用的物理渲染方程；同时用 VGGT-IR 前馈初始化替代 SfM，解决低对比度热图下位姿初始化不稳的问题。
原文依据 we explicitly model temperature, emissivity, and environmental irradiance on Gaussian primitives
摘要译文
由模型翻译 · 术语统一 · 请对照原文核对关键结论热红外（TIR）三维重建能够提供与温度场内在耦合的几何结构，即使在低光照、夜间以及烟雾遮挡的环境中也是如此。TIR 成像测量由物体温度驱动的自发热辐射，且在很大程度上不依赖外部光照；因此，若简单地将可见光谱的假设照搬到基于 TIR 的三维重建与新视角合成（NVS）中，往往会产生悬浮伪影与边缘模糊。此外，TIR 中的辐射不一致性与低对比度会削弱运动恢复结构（SfM）初始化，进而阻碍后续的 3D 高斯泼溅（3DGS）优化。我们提出 PhysIR-Splat，一个遵循红外辐射传输规律的 3DGS 框架：我们在高斯原语上显式建模温度、发射率与环境辐照度，并在渲染过程中联合考虑热发射、反射分量与大气透过率，从而生成物理一致的热红外合成结果。我们还提出 VGGT-IR，一个基于 Transformer 的前馈初始化器，它以 TIR 输入（可选加入 RGB）为输入，直接回归相机位姿与初始几何，为 PhysIR-Splat 提供模态对齐且稳定的起点。大量实验表明，我们的方法在热红外重建质量与跨视角一致性上显著超越现有方法，有效抑制了悬浮伪影并提升了边界锐度。源代码将公开于：https://github.com/JingyuanGao0919/physir-splat。
英文原文摘要 1600 字符
Thermal infrared (TIR) 3D reconstruction provides geometry that is intrinsically coupled to the temperature field, even in low-light, nighttime, and smoke-obscured environments. TIR imaging measures self-emitted thermal radiation driven by object temperature and is largely independent of external illumination; therefore, simply carrying over visible-spectrum assumptions to TIR-based 3D reconstruction and novel view synthesis (NVS) often results in floating artifacts and blurred edges. In addition, radiometric inconsistency and low contrast in TIR weaken structure-from-motion (SfM) initialization, which in turn hinders subsequent 3D Gaussian Splatting (3DGS) optimization. We present PhysIR-Splat, a 3DGS framework that follows infrared radiative transfer: we explicitly model temperature, emissivity, and environmental irradiance on Gaussian primitives and, during rendering, jointly account for thermal emission, the reflected component, and atmospheric transmittance to produce physically consistent thermal synthesis. We also introduce VGGT-IR, a Transformer-based feed-forward initializer that takes TIR input with optional RGB and directly regresses camera poses and initial geometry, providing a modality-aligned and stable starting point for PhysIR-Splat. Extensive experiments demonstrate that our method significantly surpasses existing approaches in thermal reconstruction quality and cross-view consistency, effectively suppressing floating artifacts and enhancing boundary sharpness. The source code will be publicly available at : https://github.com/JingyuanGao0919/physir-splat.