PureSample: Neural Materials Learned by Sampling Microgeometry

Abstract: Traditional physically-based material models rely on analytically derived bidirectional reflectance distribution functions (BRDFs), typically by considering statistics of micro-primitives such as facets, flakes, or spheres, sometimes combined with multi-bounce interactions such as layering and multiple scattering. These derivations are often complex and model-specific, and typically consider a statistical aggregate of a large surface area, ignoring spatial variation. Once an analytic BRDF's evaluation is defined, one still needs to design an importance sampling method for it, and a way to evaluate the pdf of that sampling distribution, requiring further model-specific derivations. We present PureSample: a novel neural BRDF representation that allows learning a material's behavior purely by sampling forward random walks on the microgeometry, which is usually straightforward to implement. Our representation allows for efficient importance sampling, pdf evaluation, and BRDF evaluation, for homogeneous as well as spatially varying materials. We achieve this by two learnable components: first, the sampling distribution is modeled using a flow matching neural network, which allows both importance sampling and pdf evaluation; second, we introduce a view-dependent albedo term, captured by a lightweight neural network, which allows for converting a scalar pdf value to a colored BRDF value for any pair of view and light directions. We demonstrate PureSample on challenging materials, including multi-layered materials, multiple-scattering microfacet materials, and various other microstructures.