Search papers, labs, and topics across Lattice.
This paper introduces STEP-Parts, a toolchain for extracting geometric instance partitions directly from STEP Boundary Representations (B-Reps) of CAD models, preserving analytic surface structure and topological adjacency. It merges adjacent B-Rep faces based on shared analytic primitive type and near-tangent continuity, creating stable partitions independent of tessellation. The toolchain processes 180,000 models in under six hours and provides instance labels and metadata for downstream learning tasks, demonstrating its utility in implicit reconstruction-segmentation and point-based backbones.
Ditch triangle meshes: STEP-Parts extracts stable, geometrically meaningful part segmentations directly from CAD B-Reps, unlocking more consistent instance-level analysis for CAD learning.
Many CAD learning pipelines discretize Boundary Representations (B-Reps) into triangle meshes, discarding analytic surface structure and topological adjacency and thereby weakening consistent instance-level analysis. We present STEP-Parts, a deterministic CAD-to-supervision toolchain that extracts geometric instance partitions directly from raw STEP B-Reps and transfers them to tessellated carriers through retained source-face correspondence, yielding instance labels and metadata for downstream learning and evaluation. The construction merges adjacent B-Rep faces only when they share the same analytic primitive type and satisfy a near-tangent continuity criterion. On ABC, same-primitive dihedral angles are strongly bimodal, yielding a threshold-insensitive low-angle regime for part extraction. Because the partition is defined on intrinsic B-Rep topology rather than on a particular triangulation, the resulting boundaries remain stable under changes in tessellation. Applied to the DeepCAD subset of ABC, the pipeline processes approximately 180{,}000 models in under six hours on a consumer CPU. We release code and precomputed labels, and show that STEP-Parts serves both as a tessellation-robust geometric reference and as a useful supervision source in two downstream probes: an implicit reconstruction--segmentation network and a dataset-level point-based backbone.
