GS-IR: 3D Gaussian Splatting for Inverse Rendering

TL;DR

GS-IR introduces a novel inverse rendering method using 3D Gaussian Splatting, enabling efficient scene reconstruction, photorealistic rendering, and relighting from multi-view images under unknown lighting.

Contribution

It extends Gaussian Splatting to inverse rendering by addressing normal estimation and occlusion modeling, achieving fast, compact, and photorealistic scene reconstruction.

Findings

Outperforms baseline methods in qualitative evaluations.

Achieves fast and compact geometry reconstruction.

Provides photorealistic novel view synthesis and relighting.

Abstract

We propose GS-IR, a novel inverse rendering approach based on 3D Gaussian Splatting (GS) that leverages forward mapping volume rendering to achieve photorealistic novel view synthesis and relighting results. Unlike previous works that use implicit neural representations and volume rendering (e.g. NeRF), which suffer from low expressive power and high computational complexity, we extend GS, a top-performance representation for novel view synthesis, to estimate scene geometry, surface material, and environment illumination from multi-view images captured under unknown lighting conditions. There are two main problems when introducing GS to inverse rendering: 1) GS does not support producing plausible normal natively; 2) forward mapping (e.g. rasterization and splatting) cannot trace the occlusion like backward mapping (e.g. ray tracing). To address these challenges, our GS-IR proposes an efficient optimization scheme that incorporates a depth-derivation-based regularization for normal estimation and a baking-based occlusion to model indirect lighting. The flexible and expressive GS representation allows us to achieve fast and compact geometry reconstruction, photorealistic novel view synthesis, and effective physically-based rendering. We demonstrate the superiority of our method over baseline methods through qualitative and quantitative evaluations on various challenging scenes.