Our new X account is live! Follow @wizwand_team for updates
WorkDL logo mark

3DGUT: Enabling Distorted Cameras and Secondary Rays in Gaussian Splatting

About

3D Gaussian Splatting (3DGS) enables efficient reconstruction and high-fidelity real-time rendering of complex scenes on consumer hardware. However, due to its rasterization-based formulation, 3DGS is constrained to ideal pinhole cameras and lacks support for secondary lighting effects. Recent methods address these limitations by tracing the particles instead, but, this comes at the cost of significantly slower rendering. In this work, we propose 3D Gaussian Unscented Transform (3DGUT), replacing the EWA splatting formulation with the Unscented Transform that approximates the particles through sigma points, which can be projected exactly under any nonlinear projection function. This modification enables trivial support of distorted cameras with time dependent effects such as rolling shutter, while retaining the efficiency of rasterization. Additionally, we align our rendering formulation with that of tracing-based methods, enabling secondary ray tracing required to represent phenomena such as reflections and refraction within the same 3D representation. The source code is available at: https://github.com/nv-tlabs/3dgrut.

Qi Wu, Janick Martinez Esturo, Ashkan Mirzaei, Nicolas Moenne-Loccoz, Zan Gojcic• 2024

Related benchmarks

TaskDatasetResultRank
Novel View SynthesisMip-NeRF 360 (test)
PSNR27.26
166
Novel View SynthesisTank & Temples (test)
PSNR22.9
23
Novel View SynthesisZipNeRF 1/8 Fisheye 1.0 (test)
PSNR24.77
8
Novel View SynthesisZipNeRF 1/4 Pinhole 1.0 (test)
PSNR25.96
8
Novel View SynthesisMipNeRF360 standard (test)
PSNR27.26
7
Novel View SynthesisMip-NeRF 360
PSNR27.74
7
Novel View SynthesisMipNeRF360 1 (test)
PSNR27.26
7
Novel View SynthesisTanks&Temples
PSNR22.86
7
Novel View SynthesisPPISP-Auto
PSNR22.05
7
Novel View SynthesisBilaRF
PSNR22.6
7
Showing 10 of 18 rows

Other info

Code

Follow for update