Differentiable Rendering of Neural SDFs through Reparameterization

TL;DR

This paper introduces a novel differentiable rendering technique for neural SDFs that accurately computes gradients for geometry optimization, enabling high-quality 3D reconstructions without segmentation masks or volumetric approximations.

Contribution

It develops a continuous warping function for neural SDFs using area-sampling and quadrature, improving gradient computation at discontinuities in differentiable rendering.

Findings

Produces comparable 3D reconstructions to recent methods

Does not require 2D segmentation masks

Avoids volumetric approximations

Abstract

We present a method to automatically compute correct gradients with respect to geometric scene parameters in neural SDF renderers. Recent physically-based differentiable rendering techniques for meshes have used edge-sampling to handle discontinuities, particularly at object silhouettes, but SDFs do not have a simple parametric form amenable to sampling. Instead, our approach builds on area-sampling techniques and develops a continuous warping function for SDFs to account for these discontinuities. Our method leverages the distance to surface encoded in an SDF and uses quadrature on sphere tracer points to compute this warping function. We further show that this can be done by subsampling the points to make the method tractable for neural SDFs. Our differentiable renderer can be used to optimize neural shapes from multi-view images and produces comparable 3D reconstructions to recent SDF-based inverse rendering methods, without the need for 2D segmentation masks to guide the geometry optimization and no volumetric approximations to the geometry.