Shape From Tracing: Towards Reconstructing 3D Object Geometry and SVBRDF Material from Images via Differentiable Path Tracing

TL;DR

This paper presents a method using differentiable path tracing to refine coarse 3D meshes and material properties from images, enabling detailed reconstructions of real-world objects despite high computational costs.

Contribution

It introduces a framework that refines initial geometry and material estimates using differentiable ray tracing, achieving high-quality reconstructions from low-resolution inputs.

Findings

High-quality reconstructions in a few hours using differentiable path tracing.

Successful disambiguation of shading, shadow, and global illumination effects.

Effective refinement of real-world object models from smartphone and 360 camera data.

Abstract

Reconstructing object geometry and material from multiple views typically requires optimization. Differentiable path tracing is an appealing framework as it can reproduce complex appearance effects. However, it is difficult to use due to high computational cost. In this paper, we explore how to use differentiable ray tracing to refine an initial coarse mesh and per-mesh-facet material representation. In simulation, we find that it is possible to reconstruct fine geometric and material detail from low resolution input views, allowing high-quality reconstructions in a few hours despite the expense of path tracing. The reconstructions successfully disambiguate shading, shadow, and global illumination effects such as diffuse interreflection from material properties. We demonstrate the impact of different geometry initializations, including space carving, multi-view stereo, and 3D neural networks. Finally, with input captured using smartphone video and a consumer 360? camera for lighting estimation, we also show how to refine initial reconstructions of real-world objects in unconstrained environments.