Split-and-Fit: Learning B-Reps via Structure-Aware Voronoi Partitioning

TL;DR

This paper presents a top-down, structure-aware neural method called NVD-Net for learning boundary representations of 3D CAD models through Voronoi partitioning, improving reconstruction quality over existing bottom-up approaches.

Contribution

The paper introduces NVD-Net, a neural network that predicts Voronoi diagrams for CAD models, enabling more accurate and structure-aware B-Rep extraction compared to prior methods.

Findings

NVD-Net effectively learns Voronoi partitions from training data.

The resulting B-Reps are more plausible and accurate.

Significant improvements in reconstruction quality over existing methods.

Abstract

We introduce a novel method for acquiring boundary representations (B-Reps) of 3D CAD models which involves a two-step process: it first applies a spatial partitioning, referred to as the ``split``, followed by a ``fit`` operation to derive a single primitive within each partition. Specifically, our partitioning aims to produce the classical Voronoi diagram of the set of ground-truth (GT) B-Rep primitives. In contrast to prior B-Rep constructions which were bottom-up, either via direct primitive fitting or point clustering, our Split-and-Fit approach is top-down and structure-aware, since a Voronoi partition explicitly reveals both the number of and the connections between the primitives. We design a neural network to predict the Voronoi diagram from an input point cloud or distance field via a binary classification. We show that our network, coined NVD-Net for neural Voronoi diagrams, can effectively learn Voronoi partitions for CAD models from training data and exhibits superior generalization capabilities. Extensive experiments and evaluation demonstrate that the resulting B-Reps, consisting of parametric surfaces, curves, and vertices, are more plausible than those obtained by existing alternatives, with significant improvements in reconstruction quality. Code will be released on https://github.com/yilinliu77/NVDNet.