Oct 2026· Proceedings of the 34th ACM International Conference on Multimedia· pp. 15705-15714· 0 citations· 57 references
Computer Science
TL;DR
GraphCAD is presented, a geometry-aware generative framework that centers modeling on the Edge-Adjacency Graph (EAG), a novel graph-based CAD representation tailored for generative tasks and introduces a geometry-aware denoising mechanism that ''renders'' node attributes into spatial point sets.
Abstract
Generative modeling for Computer-Aided Design (CAD) remains fundamentally challenged by the need to capture the intricate interplay between discrete topological connectivity and precise spatial geometry inherent in Boundary Representations (B-reps). Existing generative approaches largely formulate CAD construction as a sequence modeling problem, but consequently lack an explicit topological inductive bias. In this work, we present GraphCAD, a geometry-aware generative framework that centers modeling on the Edge-Adjacency Graph (EAG), a novel graph-based CAD representation tailored for generative tasks. GraphCAD adopts a hybrid graph diffusion formulation to jointly model the discrete and continuous components of EAG and introduces a geometry-aware denoising mechanism that ''renders'' node attributes into spatial point sets. These point sets are encoded by an SE(3)-equivariant backbone and fused into the node representations, thereby grounding the generative process in physical geometry. To support this study, we construct EAGBench, a large-scale dataset of engineering CAD models. Extensive experiments show that GraphCAD can generate complex CAD models with more than 50 faces and achieves state-of-the-art performance in geometric fidelity and distributional alignment compared with sequence-based baselines.
PolarFlow, a continuous geometric flow-matching framework based on polar decomposition and truncated spectral coordinates, achieves generative quality comparable to competitive discrete methods while substantially reducing training and sampling time.
Ya-Nan Zhao, Feng Ji, Alba Carballo-Castro et al.· 0 citations
3D scene graph generation requires understanding both object semantics and metric spatial relations. However, existing open-vocabulary methods often struggle with spatial predicates because language similarity reflects semantic co-occurrence rather than explicit geometry. We propose GHOST, a Geometry-constrained Hybrid...
Chang-Qun Feng, Wangxiandi Yin, Xin Hu et al.· Proceedings of the 34th ACM...· 0 citations
KnottedGraph is introduced, a computational framework that converts such scientific representations to knotted graphs that retain graph connectivity and spatial embedding together, enabling large-scale classification and introducing a route from topological data to LLM-assisted AI4Math discovery.
Hakan Akgun, Xian-Quan Yan, Ke-Han Liu et al.· arXiv.org· 0 citations
In this work, we introduce a new clustering method, namely T-ARC (Topology-Aware Randomized Clustering), that corrects the geometric bias of K-means by embedding topological information directly into the optimization objective. Building on the assumption that the data admits an underlying hidden structure modeled via a...
S. G. De Benedictis, A. Ang, N. Del Buono et al.· arXiv.org· 0 citations
Methods for learning continuous representations, such as latent manifolds, are fundamental and ubiquitous in modern machine learning. Yet many geometric objects combine continuous shape information with discrete combinatorial structure. This combination presents a fundamental challenge for representation learning becau...
We present novel graph-centered algorithms for the construction and representation of worldline effective field theory (WEFT) integrands that separate the topological and algebraic structure of loop integrands into independent graph-numerator pairs. In this representation, the graph encodes all propagators, worldline s...
Gregor Kälin, Jungwon Lim· 2 citations
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.