Estimating the surface of translucent objects from photometric data poses significant challenges due to complex internal light scattering. We introduce a novel method that computes a depth map from single-viewpoint photographs of a material sample, captured under multiple illuminations. Our approach leverages inverse rendering to derive a volumetric representation, including density, albedo, and phase function, from which a surface mesh is reconstructed. Beyond validation with synthetic and 3D-printed physical models, we illustrate our technique's power by successfully applying it to the digitization of fabrics, a notoriously difficult material due to its intricate translucent structure. This work advances the state-of-the-art texture stack acquisition via enhanced surface reconstruction.

Abstract Environment maps are commonly used to represent and compute far‐field illumination in virtual scenes. However, they are expensive to evaluate and sample from, limiting their applicability to real‐time rendering. Previous methods have focused on compression through spherical‐domain approximations, or on learning priors for natural, day‐light illumination. These hinder both accuracy and generality, and do not provide the probability information required for importance‐sampling Monte Carlo integration. We propose NEnv, a deep‐learning fully‐differentiable method, capable of compressing and learning to sample from a single environment map. NEnv is composed of two different neural networks: A normalizing flow, able to map samples from uniform distributions to the probability density of the illumination, also providing their corresponding probabilities; and an implicit neural representation which compresses the environment map into an efficient differentiable function. The computation time of environment samples with NEnv is two orders of magnitude less than with traditional methods. NEnv makes no assumptions regarding the content (i.e. natural illumination), thus achieving higher generality than previous learning‐based approaches. We share our implementation and a diverse dataset of trained neural environment maps, which can be easily integrated into existing rendering engines.