Target-Guided Bayesian Flow Networks for Quantitatively Constrained CAD Generation
This work addresses the challenge of multi-modal data generation for CAD design, which is incremental as it builds on existing generative models but introduces a unified approach for discrete and continuous parameters.
The paper tackles the problem of generating parametric CAD sequences under quantitative constraints by proposing a novel framework called Target-Guided Bayesian Flow Network (TGBFN), which achieves state-of-the-art performance in high-fidelity, condition-aware generation tasks.
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.