用流模型并行生成高质量3D网格，速度提升18倍
MeshFlow: Efficient Artistic Mesh Generation via MeshVAE and Flow-based Diffusion Transformer
- 先用VAE将网格顶点与连接关系编码到连续潜空间
- 基于流变换器并行生成全部顶点和边，速度比自回归方法快18倍
- 适合需要快速生成艺术级3D网格的设计师与开发者
我们提出MeshFlow，一种生成类艺术家风格3D网格的新方法。现有网格生成器多采用自回归（AR）逐项预测，虽契合网格拓扑的离散特性，但推理成本随网格规模呈二次增长，且需对顶点坐标离散化，引入量化误差。为此，我们设计了一种变分自编码器（VAE），在对比损失监督下将连续顶点位置与离散连接关系统一表示于连续潜空间，其潜空间显著优于以往基于标记的网格表示。在此基础上，构建基于修正流（Rectified Flow）的3D生成器，可并行生成所有网格顶点与边。实验表明，该模型生成速度比最快自回归生成器快18倍，同时在标准网格生成指标上保持优异性能。
原文摘要 · Abstract (English)
We present MeshFlow, a new method for generating artist-like 3D meshes. Current mesh generators often adopt Auto-Regressive (AR) next-token prediction, a natural choice given the discrete nature of mesh topology. However, AR methods scale poorly because the inference cost is quadratic in mesh size. They also require discretizing the vertex coordinates, which introduces quantization errors. To address these challenges, we introduce a Variational Autoencoder (VAE) that, supervised with a contrastive loss, represents both continuous vertex positions and discrete connectivity in a continuous latent space. This latent space is significantly more compact than prior token-based mesh representations. We then build a 3D generator based on a Rectified Flow transformer, generating all mesh vertices and edges in parallel. Our model generates meshes 18x faster than the fastest AR generator while also achieving excellent accuracy across standard mesh-generation metrics. Homepage: https://mesh-flow.github.io/, Code: https://github.com/facebookresearch/meshflow
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。