On Neural BRDFs: A Thorough Comparison of State-of-the-Art Approaches

TL;DR

This paper provides a comprehensive comparison of neural BRDF modeling approaches, evaluating their qualitative and quantitative performance, and introduces two extensions to improve existing methods by ensuring reciprocity and splitting reflectance.

Contribution

It offers the first thorough evaluation of neural BRDF methods and proposes two novel extensions for improved physical plausibility and reflectance modeling.

Findings

Neural BRDF approaches vary significantly in reconstruction quality.

The proposed additive strategy effectively separates diffuse and specular components.

Input mapping ensures reciprocity exactly, outperforming soft constraint methods.

Abstract

The bidirectional reflectance distribution function (BRDF) is an essential tool to capture the complex interaction of light and matter. Recently, several works have employed neural methods for BRDF modeling, following various strategies, ranging from utilizing existing parametric models to purely neural parametrizations. While all methods yield impressive results, a comprehensive comparison of the different approaches is missing in the literature. In this work, we present a thorough evaluation of several approaches, including results for qualitative and quantitative reconstruction quality and an analysis of reciprocity and energy conservation. Moreover, we propose two extensions that can be added to existing approaches: A novel additive combination strategy for neural BRDFs that split the reflectance into a diffuse and a specular part, and an input mapping that ensures reciprocity exactly by construction, while previous approaches only ensure it by soft constraints.