Rethinking Directional Parameterization in Neural Implicit Surface Reconstruction

TL;DR

This paper introduces a hybrid directional parameterization for neural implicit surface reconstruction, improving the modeling of complex and specular surfaces by overcoming limitations of traditional directional representations.

Contribution

It proposes a novel hybrid directional parameterization that enhances reconstruction quality for diverse materials and geometries, and is easily integrated into existing methods.

Findings

Consistently improves reconstruction of specular and complex surfaces.

Outperforms traditional directional parameterizations in diverse scenarios.

Nearly parameter-free and easy to adopt in current frameworks.

Abstract

Multi-view 3D surface reconstruction using neural implicit representations has made notable progress by modeling the geometry and view-dependent radiance fields within a unified framework. However, their effectiveness in reconstructing objects with specular or complex surfaces is typically biased by the directional parameterization used in their view-dependent radiance network. {\it Viewing direction} and {\it reflection direction} are the two most commonly used directional parameterizations but have their own limitations. Typically, utilizing the viewing direction usually struggles to correctly decouple the geometry and appearance of objects with highly specular surfaces, while using the reflection direction tends to yield overly smooth reconstructions for concave or complex structures. In this paper, we analyze their failed cases in detail and propose a novel hybrid directional parameterization to address their limitations in a unified form. Extensive experiments demonstrate the proposed hybrid directional parameterization consistently delivered satisfactory results in reconstructing objects with a wide variety of materials, geometry and appearance, whereas using other directional parameterizations faces challenges in reconstructing certain objects. Moreover, the proposed hybrid directional parameterization is nearly parameter-free and can be effortlessly applied in any existing neural surface reconstruction method.