Parameterization-driven Neural Surface Reconstruction for Object-oriented Editing in Neural Rendering

TL;DR

This paper presents a neural surface parameterization method that simplifies neural implicit surfaces into easy-to-edit parametric domains, enabling intuitive object editing in neural rendering without prior info.

Contribution

It introduces a novel neural algorithm for parameterizing implicit surfaces to simple domains like spheres and polycubes, facilitating intuitive editing and automatic texture mapping.

Findings

Effective parameterization of neural implicit surfaces achieved

Automatic texture map computation demonstrated

Applicable to human heads and man-made objects

Abstract

The advancements in neural rendering have increased the need for techniques that enable intuitive editing of 3D objects represented as neural implicit surfaces. This paper introduces a novel neural algorithm for parameterizing neural implicit surfaces to simple parametric domains like spheres and polycubes. Our method allows users to specify the number of cubes in the parametric domain, learning a configuration that closely resembles the target 3D object's geometry. It computes bi-directional deformation between the object and the domain using a forward mapping from the object's zero level set and an inverse deformation for backward mapping. We ensure nearly bijective mapping with a cycle loss and optimize deformation smoothness. The parameterization quality, assessed by angle and area distortions, is guaranteed using a Laplacian regularizer and an optimized learned parametric domain. Our framework integrates with existing neural rendering pipelines, using multi-view images of a single object or multiple objects of similar geometries to reconstruct 3D geometry and compute texture maps automatically, eliminating the need for any prior information. We demonstrate the method's effectiveness on images of human heads and man-made objects.