Extracting Triangular 3D Models, Materials, and Lighting From Images

TL;DR

This paper introduces a method for extracting detailed 3D triangle meshes, materials, and lighting from images, enabling realistic scene editing and rendering in traditional graphics engines.

Contribution

It presents a novel joint optimization approach that produces deployable triangle meshes with spatially-varying materials and environment lighting, using differentiable rendering techniques.

Findings

Models enable advanced scene editing and material decomposition.

Achieves high-quality view interpolation at interactive rates.

Outputs are compatible with standard graphics engines.

Abstract

We present an efficient method for joint optimization of topology, materials and lighting from multi-view image observations. Unlike recent multi-view reconstruction approaches, which typically produce entangled 3D representations encoded in neural networks, we output triangle meshes with spatially-varying materials and environment lighting that can be deployed in any traditional graphics engine unmodified. We leverage recent work in differentiable rendering, coordinate-based networks to compactly represent volumetric texturing, alongside differentiable marching tetrahedrons to enable gradient-based optimization directly on the surface mesh. Finally, we introduce a differentiable formulation of the split sum approximation of environment lighting to efficiently recover all-frequency lighting. Experiments show our extracted models used in advanced scene editing, material decomposition, and high quality view interpolation, all running at interactive rates in triangle-based renderers (rasterizers and path tracers). Project website: https://nvlabs.github.io/nvdiffrec/ .