Target-Guided Bayesian Flow Networks for Quantitatively Constrained CAD Generation

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Zheng, Wenhao, Sun, Chenwei, Zhang, Wenbo, Lv, Jiancheng, Liu, Xianggen
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866918177695334400
author Zheng, Wenhao
Sun, Chenwei
Zhang, Wenbo
Lv, Jiancheng
Liu, Xianggen
author_facet Zheng, Wenhao
Sun, Chenwei
Zhang, Wenbo
Lv, Jiancheng
Liu, Xianggen
contents Deep generative models, such as diffusion models, have shown promising progress in image generation and audio generation via simplified continuity assumptions. However, the development of generative modeling techniques for generating multi-modal data, such as parametric CAD sequences, still lags behind due to the challenges in addressing long-range constraints and parameter sensitivity. In this work, we propose a novel framework for quantitatively constrained CAD generation, termed Target-Guided Bayesian Flow Network (TGBFN). For the first time, TGBFN handles the multi-modality of CAD sequences (i.e., discrete commands and continuous parameters) in a unified continuous and differentiable parameter space rather than in the discrete data space. In addition, TGBFN penetrates the parameter update kernel and introduces a guided Bayesian flow to control the CAD properties. To evaluate TGBFN, we construct a new dataset for quantitatively constrained CAD generation. Extensive comparisons across single-condition and multi-condition constrained generation tasks demonstrate that TGBFN achieves state-of-the-art performance in generating high-fidelity, condition-aware CAD sequences. The code is available at https://github.com/scu-zwh/TGBFN.
format Preprint
id arxiv_https___arxiv_org_abs_2510_25163
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Target-Guided Bayesian Flow Networks for Quantitatively Constrained CAD Generation
Zheng, Wenhao
Sun, Chenwei
Zhang, Wenbo
Lv, Jiancheng
Liu, Xianggen
Computer Vision and Pattern Recognition
Deep generative models, such as diffusion models, have shown promising progress in image generation and audio generation via simplified continuity assumptions. However, the development of generative modeling techniques for generating multi-modal data, such as parametric CAD sequences, still lags behind due to the challenges in addressing long-range constraints and parameter sensitivity. In this work, we propose a novel framework for quantitatively constrained CAD generation, termed Target-Guided Bayesian Flow Network (TGBFN). For the first time, TGBFN handles the multi-modality of CAD sequences (i.e., discrete commands and continuous parameters) in a unified continuous and differentiable parameter space rather than in the discrete data space. In addition, TGBFN penetrates the parameter update kernel and introduces a guided Bayesian flow to control the CAD properties. To evaluate TGBFN, we construct a new dataset for quantitatively constrained CAD generation. Extensive comparisons across single-condition and multi-condition constrained generation tasks demonstrate that TGBFN achieves state-of-the-art performance in generating high-fidelity, condition-aware CAD sequences. The code is available at https://github.com/scu-zwh/TGBFN.
title Target-Guided Bayesian Flow Networks for Quantitatively Constrained CAD Generation
topic Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2510.25163