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

TL;DR

Brep2Shape introduces a self-supervised transformer-based approach to align boundary and shape representations in CAD models, bridging the gap between abstract and intuitive geometries for improved accuracy and efficiency.

Contribution

The paper presents a novel self-supervised pre-training method with a dual transformer architecture and topology attention to better align B-rep and shape representations in CAD.

Findings

Achieves state-of-the-art accuracy in shape representation tasks.

Demonstrates faster convergence compared to existing methods.

Offers scalable solutions for CAD boundary and shape alignment.

Abstract

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 B\'ezier 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.