An adaptive parameterization for efficient material acquisition and rendering

Abstract

One of the key ingredients of any physically based rendering system is a detailed specification characterizing the interaction of light and matter of all materials present in a scene, typically via the Bidirectional Reflectance Distribution Function (BRDF). Despite their utility, access to real-world BRDF datasets remains limited: this is because measurements involve scanning a four-dimensional domain at sufficient resolution, a tedious and often infeasibly time-consuming process. We propose a new parameterization that automatically adapts to the behavior of a material, warping the underlying 4D domain so that most of the volume maps to regions where the BRDF takes on non-negligible values, while irrelevant regions are strongly compressed. This adaptation only requires a brief 1D or 2D measurement of the material's retro-reflective properties. Our parameterization is unified in the sense that it combines several steps that previously required intermediate data conversions: the same mapping can simultaneously be used for BRDF acquisition, storage, and it supports efficient Monte Carlo sample generation. We observe that the above desiderata are satisfied by a core operation present in modern rendering systems, which maps uniform variates to direction samples that are proportional to an analytic BRDF. Based on this insight, we define our adaptive parameterization as an invertible, retro-reflectively driven mapping between the parametric and directional domains. We are able to create noise-free renderings of existing BRDF datasets after conversion into our representation with the added benefit that the warped data is significantly more compact, requiring 16KiB and 544KiB per spectral channel for isotropic and anisotropic specimens, respectively. Finally, we show how to modify an existing gonio-photometer to provide the needed retro-reflection measurements. Acquisition then proceeds within a 4D space that is warped by our parameterization. We demonstrate the efficacy of this scheme by acquiring the first set of spectral BRDFs of surfaces exhibiting arbitrary roughness, including anisotropy.

Three Perceptual Dimensions for Specular and Diffuse Reflection

“…Please note that, once itted to the ABC model, the appearance of some measured materials in the MERL dataset might not be faithfully reproduced [46]; similar considerations apply to any current analytical BRDF model [37]. Several works have pointed out a number of measurement artifacts in the MERL database, such as optical aberrations, discontinuities, and extrapolation artifacts for data above 75 degrees [24,45,66], which cause the measured materials in the MERL dataset to difer from the physical counterpart. Such artifacts explain in part the diiculty to faithfully reproduce the appearance of some materials in the MERL dataset, and highlight the importance of using a gamut beyond the one deined by the dataset itself.…”

“…Apart from color, most of these models separate relectance into difuse (following Lambert's law [56]) and specular dimensions and have parameters controlling the their angular distributions [14,29,37]. Alternative approaches use measured BRDF data, for instance from existing datasets such as MERL [64], UTIA [27] and RGL [23], to convert BRDFs into products of compact non-parametric lower-dimensional factors [6], or to derive low-dimensional embeddings, in which the manifold representation can be used for rendering [89].…”

A similarity measure for material appearance

“…Our model could potentially be used as a feature extractor, or as a baseline for transferlearning [Sharif Razavian et al 2014;Yosinski et al 2014] in other material perception tasks. A larger database could translate into an improvement of our model's predictions; upcoming databases of complex measured materials (e.g., Dupuy et al [2018]) could be used to expand our training data and lead to a richer and more accurate analysis of appearance. Our methodology for data collection and model training could be useful in these cases.…”

BxDF material acquisition, representation, and rendering for VR and design

“…Backscattering data can be used to extract an appropriate distribution for microfacets BRDF models [Ashikhmin and Premoze 2007;Dupuy and Jakob 2018]. Based on this observation, mobile devices equipped with a lash light, typically positioned near the back camera, represent near-coaxial setups particularly useful to capture the backscatter surface relectance to be itted in a microfacets BRDF model [Riviere et al 2015], since they allow to extract an appropriate distribution of microfacets [Ashikhmin and Premoze 2007], by observing high frequencies for specular components in the highlights areas and the difuse components in other areas .…”