HiDiGen: Hierarchical Diffusion for B-Rep Generation with Explicit Topological Constraints

arXiv cs.CV / 4/6/2026

💬 OpinionSignals & Early TrendsIdeas & Deep AnalysisModels & Research

共有:

Key Points

The paper introduces HiDiGen, a hierarchical diffusion framework aimed at generating valid CAD boundary representations (B-reps) by addressing the difficulty of jointly modeling discrete topology and continuous geometry.
HiDiGen decouples the task into two stages: first building an explicit topological scaffold via face–edge incidence relations, then generating and refining geometric elements with Transformer-based diffusion modules.
During geometry refinement, the method dynamically establishes and enforces adjacency constraints (e.g., edge–vertex relationships) to maintain structural consistency and improve validity.
The authors report that the two-stage topological modeling and progressive geometry hierarchy produce novel, diverse, and topologically sound CAD models, with strong experimental performance.

Abstract

Boundary representation (B-rep) is the standard 3D modeling format in CAD systems, encoding both geometric primitives and topological connectivity. Despite its prevalence, deep generative modeling of valid B-rep structures remains challenging due to the intricate interplay between discrete topology and continuous geometry. In this paper, we propose HiDiGen, a hierarchical generation framework that decouples geometry modeling into two stages, each guided by explicitly modeled topological constraints. Specifically, our approach first establishes face-edge incidence relations to define a coherent topological scaffold, upon which face proxies and initial edge curves are generated. Subsequently, multiple Transformer-based diffusion modules are employed to refine the geometry by generating precise face surfaces and vertex positions, with edge-vertex adjacencies dynamically established and enforced to preserve structural consistency. This progressive geometry hierarchy enables the generation of more novel and diverse shapes, while two-stage topological modeling ensures high validity. Experimental results show that HiDiGen achieves strong performance, generating novel, diverse, and topologically sound CAD models.