Neural BRDFs: Representation and Operations

TL;DR

This paper introduces a neural network-based framework for representing and manipulating BRDFs in a flexible, efficient latent space, enabling operations like layering and interpolation without fixed, pre-computed functions.

Contribution

It presents a novel neural BRDF representation that allows for flexible operations directly in latent space, enhancing the utility of neural BRDF models.

Findings

Accurate BRDF compression into latent vectors.

Enables operations like layering and interpolation in latent space.

Efficient evaluation and sampling compared to Monte Carlo methods.

Abstract

Bidirectional reflectance distribution functions (BRDFs) are pervasively used in computer graphics to produce realistic physically-based appearance. In recent years, several works explored using neural networks to represent BRDFs, taking advantage of neural networks' high compression rate and their ability to fit highly complex functions. However, once represented, the BRDFs will be fixed and therefore lack flexibility to take part in follow-up operations. In this paper, we present a form of "Neural BRDF algebra", and focus on both representation and operations of BRDFs at the same time. We propose a representation neural network to compress BRDFs into latent vectors, which is able to represent BRDFs accurately. We further propose several operations that can be applied solely in the latent space, such as layering and interpolation. Spatial variation is straightforward to achieve by using textures of latent vectors. Furthermore, our representation can be efficiently evaluated and sampled, providing a competitive solution to more expensive Monte Carlo layering approaches.