发表机构
The University of Osaka; Osaka Electro-Communication University（大阪大学; 大阪电气通信大学）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
OptimusMesh提出一种从点云直接生成紧凑三角网格的自回归框架，通过将2048个点压缩为16个稀疏潜在枢轴，显著减少条件序列，以更少的面数实现高保真几何输出。
AI 中文摘要
直接从点云生成紧凑且几何保真的三维网格仍然是一个基本挑战。点云是无序且稀疏的，而网格则表现出不规则的结构和变化的拓扑。因此，许多现有方法依赖于隐式表示，随后进行表面提取或重建。尽管这些方法有效，但它们可能产生密集或过度平滑的网格，通常需要计算昂贵的后处理和简化。我们提出了OptimusMesh，一个利用稀疏潜在枢轴条件化直接从点云生成紧凑三角形网格的框架。我们的关键思想是将$2{,}048$个有向输入点压缩为仅$16$个稀疏潜在枢轴，将几何条件集减少$128$倍。这些枢轴提供了一个紧凑的结构表示，共享于一个两阶段自回归框架中，该框架首先生成网格顶点，然后基于生成的顶点和相同的枢轴预测三角形面。与评估的近期点云条件自回归方法（使用$257$个解码器条件令牌）相比，OptimusMesh仅使用$16$个，产生$16.1$倍更短的条件序列。实验表明，OptimusMesh在比较的近期自回归方法中产生最紧凑的输出，使用$25.7\%$--$94.1\%$更少的面，同时保持竞争力的几何保真度和分布质量。
英文摘要
Generating compact and geometrically faithful 3D meshes directly from point clouds remains a fundamental challenge. Point clouds are unordered and sparse, whereas meshes exhibit irregular structure and varying topology. As a result, many existing approaches rely on implicit representations followed by surface extraction or reconstruction. Although effective, these pipelines can produce dense or over-smoothed meshes, often requiring computationally expensive post-processing and simplification. We present OptimusMesh, a framework for direct compact triangle mesh generation from point clouds using sparse latent pivot conditioning. Our key idea is to compress $2{,}048$ oriented input points into only $16$ sparse latent pivots, reducing the geometric conditioning set by $128\times$. These pivots provide a compact structural representation shared across a two-stage autoregressive framework that first generates mesh vertices and then predicts triangular faces conditioned on the generated vertices and the same pivots. Compared with the evaluated recent point-cloud-conditioned autoregressive methods, which use $257$ decoder-conditioning tokens, OptimusMesh uses only $16$, yielding a $16.1\times$ shorter conditioning sequence. Experiments show that OptimusMesh produces the most compact outputs among the compared recent autoregressive methods, using $25.7\%$--$94.1\%$ fewer faces while maintaining competitive geometric fidelity and distributional quality.
