TextureSplat: Per-Primitive Texture Mapping for Reflective Gaussian Splatting

TL;DR

TextureSplat introduces a novel per-primitive texture mapping approach for Gaussian Splatting that enhances the modeling of complex, reflective surfaces by incorporating spatially variable normals and material properties, leveraging GPU acceleration.

Contribution

It presents a physically grounded Gaussian Splatting method with per-primitive textures to better capture complex surface interactions, especially in reflective scenes.

Findings

Improved rendering of reflective surfaces with high-frequency specular components.

Enhanced GPU-based rendering speed using unified material texture atlas.

Demonstrated effectiveness on complex scene scenarios.

Abstract

Gaussian Splatting have demonstrated remarkable novel view synthesis performance at high rendering frame rates. Optimization-based inverse rendering within complex capture scenarios remains however a challenging problem. A particular case is modelling complex surface light interactions for highly reflective scenes, which results in intricate high frequency specular radiance components. We hypothesize that such challenging settings can benefit from increased representation power. We hence propose a method that tackles this issue through a geometrically and physically grounded Gaussian Splatting borne radiance field, where normals and material properties are spatially variable in the primitive's local space. Using per-primitive texture maps for this purpose, we also propose to harness the GPU hardware to accelerate rendering at test time via unified material texture atlas. Code will be available at https://github.com/maeyounes/TextureSplat