High-fidelity 3D assets with materials composed of fibers (including hair),complex layered material shaders, or fine scattering geometry are criticalin high-end realistic rendering applications. Rendering such models is computationally expensive due to heavy shaders and long scattering paths.Moreover, implementing the shading and scattering models is non-trivialand has to be done not only in the 3D content authoring software (whichis necessarily complex), but also in all downstream rendering solutions.For example, web and mobile viewers for complex 3D assets are desirable,but frequently cannot support the full shading complexity allowed by theauthoring application. Our goal is to design a neural representation for 3Dassets with complex shading that supports full relightability and full integrationinto existing renderers.We provide an end-to-end shading solutionat the first intersection of a ray with the underlying geometry. All shadingand scattering is precomputed and included in the neural asset; no multiplescattering paths need to be traced, and no complex shading models need tobe implemented to render our assets, beyond a single neural architecture.We combine an MLP decoder with a feature grid. Shading consists ofquerying a feature vector, followed by an MLP evaluation producing thefinal reflectance value. Our method provides high-fidelity shading, closeto the ground-truth Monte Carlo estimate even at close-up views. Webelieve our neural assets could be used in practical renderers, providingsignificant speed-ups and simplifying renderer implementations.
NETL
NSTL
Assoc Computing Machinery
