Brep2Shape: Boundary and Shape Representation Alignment via Self-Supervised Transformers

Fuente: arXiv
Enregistré dans:
Détails bibliographiques
Auteurs principaux: Sun, Yuanxu, Ma, Yuezhou, Wu, Haixu, Zeng, Guanyang, Chen, Muye, Wang, Jianmin, Long, Mingsheng
Format: Preprint
Publié: 2026
Sujets:
Accès en ligne:
Tags: Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
_version_ 1866911430853263360
author Sun, Yuanxu
Ma, Yuezhou
Wu, Haixu
Zeng, Guanyang
Chen, Muye
Wang, Jianmin
Long, Mingsheng
author_facet Sun, Yuanxu
Ma, Yuezhou
Wu, Haixu
Zeng, Guanyang
Chen, Muye
Wang, Jianmin
Long, Mingsheng
contents Boundary representation (B-rep) is the industry standard for computer-aided design (CAD). While deep learning shows promise in processing B-rep models, existing methods suffer from a representation gap: continuous approaches offer analytical precision but are visually abstract, whereas discrete methods provide intuitive clarity at the expense of geometric precision. To bridge this gap, we introduce Brep2Shape, a novel self-supervised pre-training method designed to align abstract boundary representations with intuitive shape representations. Our method employs a geometry-aware task where the model learns to predict dense spatial points from parametric Bézier control points, enabling the network to better understand physical manifolds derived from abstract coefficients. To enhance this alignment, we propose a Dual Transformer backbone with parallel streams that independently encode surface and curve tokens to capture their distinct geometric properties. Moreover, the topology attention is integrated to model the interdependencies between surfaces and curves, thereby maintaining topological consistency. Experimental results demonstrate that Brep2Shape offers significant scalability, achieving state-of-the-art accuracy and faster convergence across various downstream tasks.
format Preprint
id arxiv_https___arxiv_org_abs_2602_07429
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Brep2Shape: Boundary and Shape Representation Alignment via Self-Supervised Transformers
Sun, Yuanxu
Ma, Yuezhou
Wu, Haixu
Zeng, Guanyang
Chen, Muye
Wang, Jianmin
Long, Mingsheng
Machine Learning
Artificial Intelligence
Boundary representation (B-rep) is the industry standard for computer-aided design (CAD). While deep learning shows promise in processing B-rep models, existing methods suffer from a representation gap: continuous approaches offer analytical precision but are visually abstract, whereas discrete methods provide intuitive clarity at the expense of geometric precision. To bridge this gap, we introduce Brep2Shape, a novel self-supervised pre-training method designed to align abstract boundary representations with intuitive shape representations. Our method employs a geometry-aware task where the model learns to predict dense spatial points from parametric Bézier control points, enabling the network to better understand physical manifolds derived from abstract coefficients. To enhance this alignment, we propose a Dual Transformer backbone with parallel streams that independently encode surface and curve tokens to capture their distinct geometric properties. Moreover, the topology attention is integrated to model the interdependencies between surfaces and curves, thereby maintaining topological consistency. Experimental results demonstrate that Brep2Shape offers significant scalability, achieving state-of-the-art accuracy and faster convergence across various downstream tasks.
title Brep2Shape: Boundary and Shape Representation Alignment via Self-Supervised Transformers
topic Machine Learning
Artificial Intelligence
url https://arxiv.org/abs/2602.07429