From Transparent to Opaque: Rethinking Neural Implicit Surfaces with $\alpha$-NeuS

TL;DR

This paper introduces $eta$-NeuS, an extension of NeuS, which effectively reconstructs both transparent and opaque 3D objects by analyzing the learned distance field and developing a new surface extraction method, validated on a new benchmark.

Contribution

The paper presents $eta$-NeuS, a novel method that accurately reconstructs transparent and opaque objects simultaneously, addressing limitations of previous neural surface methods.

Findings

NeuS is unbiased for materials from transparent to opaque.

Transparent and opaque surfaces correspond to different minima in the distance field.

The proposed extraction method effectively reconstructs mixed-material scenes.

Abstract

Traditional 3D shape reconstruction techniques from multi-view images, such as structure from motion and multi-view stereo, face challenges in reconstructing transparent objects. Recent advances in neural radiance fields and its variants primarily address opaque or transparent objects, encountering difficulties to reconstruct both transparent and opaque objects simultaneously. This paper introduces $\alpha$-Neus -- an extension of NeuS -- that proves NeuS is unbiased for materials from fully transparent to fully opaque. We find that transparent and opaque surfaces align with the non-negative local minima and the zero iso-surface, respectively, in the learned distance field of NeuS. Traditional iso-surfacing extraction algorithms, such as marching cubes, which rely on fixed iso-values, are ill-suited for such data. We develop a method to extract the transparent and opaque surface simultaneously based on DCUDF. To validate our approach, we construct a benchmark that includes both real-world and synthetic scenes, demonstrating its practical utility and effectiveness. Our data and code are publicly available at https://github.com/728388808/alpha-NeuS.