Scheduling the Off-Diagonal Weingarten Loss of Neural SDFs for CAD Models

TL;DR

This paper introduces scheduling strategies for the Off-Diagonal Weingarten loss in neural SDFs, improving CAD model reconstruction by dynamically adjusting regularization strength during training.

Contribution

It proposes time-varying scheduling methods for the ODW loss, enhancing reconstruction accuracy over fixed regularization weights.

Findings

Time-varying schedules outperform fixed weights in experiments.

Up to 35% improvement in Chamfer Distance over baseline.

Scheduling enables better balance between stabilization and detail recovery.

Abstract

Neural signed distance functions (SDFs) have become a powerful representation for geometric reconstruction from point clouds, yet they often require both gradient- and curvature-based regularization to suppress spurious warp and preserve structural fidelity. FlatCAD introduced the Off-Diagonal Weingarten (ODW) loss as an efficient second-order prior for CAD surfaces, approximating full-Hessian regularization at roughly half the computational cost. However, FlatCAD applies a fixed ODW weight throughout training, which is suboptimal: strong regularization stabilizes early optimization but suppresses detail recovery in later stages. We present scheduling strategies for the ODW loss that assign a high initial weight to stabilize optimization and progressively decay it to permit fine-scale refinement. We investigate constant, linear, quintic, and step interpolation schedules, as well as an increasing warm-up variant. Experiments on the ABC CAD dataset demonstrate that time-varying schedules consistently outperform fixed weights. Our method achieves up to a 35% improvement in Chamfer Distance over the FlatCAD baseline, establishing scheduling as a simple yet effective extension of curvature regularization for robust CAD reconstruction.