FeatureFox: Sample-Efficient Panoptic Graph Segmentation for Machining Feature Recognition in B-Rep 3D-CAD Models

TL;DR

FeatureFox is a sample-efficient panoptic graph segmentation pipeline for machining feature recognition in B-Rep 3D-CAD models, outperforming deep learning baselines in data efficiency and practical applicability.

Contribution

We introduce FeatureFox, a novel, efficient pipeline for joint instance segmentation and semantic labeling in CAD models, reducing data and compute requirements significantly.

Findings

FeatureFox achieves PQ > 0.9 with ~250 training parts, compared to ~5,000 for AAGNet.

FeatureFox trains in seconds on a GPU, demonstrating high efficiency.

It generalizes well to unseen industrial CAD parts, showing practical potential.

Abstract

Automatic feature recognition (AFR) on B-Rep 3D-CAD models is central to CAD/CAM automation, yet most learning-based methods are complex, data-hungry, and evaluate instance grouping and semantic labeling separately. We present FeatureFox, a panoptic AFR pipeline that outputs machining instances with semantic labels: a calibrated binary edge classifier on enriched edge attributes localizes feature boundaries, instances are recovered as connected components in a pruned face-adjacency graph, and a per-instance classifier predicts the machining class from aggregated subgraph attributes. We evaluate on MFInstSeg using Panoptic Quality (PQ), which jointly scores instance separation and semantic correctness. FeatureFox is substantially more sample- and compute-efficient than the deep baseline AAGNet, reaching $\mathrm{PQ}>0.9$ with $\sim250$ training parts versus $\sim5{,}000$ for AAGNet, and training on the full MFInstSeg set takes seconds on a GPU. On the full training set, AAGNet surpasses FeatureFox marginally in PQ, while FeatureFox remains slightly ahead in feature-level recognition and localization accuracy. Finally, leveraging its low data requirement, we train FeatureFox on $270$ manually labeled industrial CAD parts and show qualitative generalization to an unseen real industrial part, indicating practical real-world applicability.