HoLa: B-Rep Generation using a Holistic Latent Representation

2025-04-19

AI Summary

This work addresses geometric ambiguity, topological inconsistency, and structural redundancy in CAD B-Rep modeling. We propose HoLa (Holographic Latent Space), a unified representation framework that reformulates B-Rep topology learning as a continuous geometric reconstruction task—specifically, reconstructing surface intersection curves—using surfaces alone as latent carriers to jointly encode faces, edges, vertices, and their connectivity. This marks the first approach to cast B-Rep topology learning into Euclidean-space-based differentiable geometry reconstruction. Furthermore, we introduce the first diffusion-based B-Rep generator supporting multimodal inputs—including point clouds, images, sketches, and text—while eliminating conventional multi-stage pipelines. Experiments demonstrate an 82% generation accuracy (surpassing prior SOTA of ~50%), significantly mitigating geometric ambiguity, structural redundancy, and topological mismatch, alongside substantially reduced training complexity.

Technology Category

• Computer Vision: Diffusion Models for Vision
• Machine Learning: Deep Generative Models & Autoencoders
• Knowledge Representation and Reasoning: Geometric, Spatial, and Temporal Reasoning

Application Category

• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs
• Economics, Online Markets and Human Computation: Economic ramifications for generative AI infrastructure and applications
• Semantics and Knowledge: Data modeling to support human-machine intelligence, including LLMs agents, intelligent system behavior, explanations, and user-friendly interactions

Abstract

We introduce a novel representation for learning and generating Computer-Aided Design (CAD) models in the form of $ extit{boundary representations}$ (B-Reps). Our representation unifies the continuous geometric properties of B-Rep primitives in different orders (e.g., surfaces and curves) and their discrete topological relations in a $ extit{holistic latent}$ (HoLa) space. This is based on the simple observation that the topological connection between two surfaces is intrinsically tied to the geometry of their intersecting curve. Such a prior allows us to reformulate topology learning in B-Reps as a geometric reconstruction problem in Euclidean space. Specifically, we eliminate the presence of curves, vertices, and all the topological connections in the latent space by learning to distinguish and derive curve geometries from a pair of surface primitives via a neural intersection network. To this end, our holistic latent space is only defined on surfaces but encodes a full B-Rep model, including the geometry of surfaces, curves, vertices, and their topological relations. Our compact and holistic latent space facilitates the design of a first diffusion-based generator to take on a large variety of inputs including point clouds, single/multi-view images, 2D sketches, and text prompts. Our method significantly reduces ambiguities, redundancies, and incoherences among the generated B-Rep primitives, as well as training complexities inherent in prior multi-step B-Rep learning pipelines, while achieving greatly improved validity rate over current state of the art: 82% vs. $approx$50%.

Problem

• Unify geometric and topological CAD properties in latent space
• Reformulate topology learning as geometric reconstruction
• Generate valid B-Reps from diverse inputs via diffusion

Innovation

• Holistic latent space unifies geometry and topology
• Neural intersection network derives curve geometries
• Diffusion-based generator handles diverse input types