MeshRipple：艺术家网格的结构化自回归生成
MeshRipple: Structured Autoregressive Generation of Artist-Meshes
把网格面序列化顺序与表面拓扑对齐（frontier-aware BFS tokenization），并用稀疏注意力全局记忆扩展感受野，可直接迁移到任何自回归式 3D 资产生成或点云/网格序列建模中，解决滑窗推理割裂长程几何依赖的问题。
原文依据 a sparse-attention global memory that provides an effectively unbounded receptive field to resolve long-range topological dependencies
摘要译文
由模型翻译 · 术语统一 · 请对照原文核对关键结论网格是三维资产的主要表示形式。自回归网格生成器将面序列化为序列，并在截断片段上训练、配合滑动窗口推理以应对内存限制。然而，这种训练与推理的不匹配破坏了长程几何依赖关系，产生孔洞与碎片化部件。为解决这一关键局限，我们提出 MeshRipple，它从活跃的生成前沿向外扩展网格，如同表面上的涟漪。MeshRipple 基于三项关键创新：一种前沿感知的 BFS 词元化方法，使生成顺序与表面拓扑对齐；一种扩张式预测策略，维持连贯且连通的表面生长；以及一种稀疏注意力全局记忆，提供实际上无界的感受野以解决长程拓扑依赖。这一集成设计使 MeshRipple 能够生成具有高表面保真度与拓扑完整性的网格，优于近期强有力的基线方法。
英文原文摘要 987 字符
Meshes serve as a primary representation for 3D assets. Autoregressive mesh generators serialize faces into sequences and train on truncated segments with sliding-window inference to cope with memory limits. However, this mismatch breaks long-range geometric dependencies, producing holes and fragmented components. To address this critical limitation, we introduce MeshRipple, which expands a mesh outward from an active generation frontier, akin to a ripple on a surface. MeshRipple rests on three key innovations: a frontier-aware BFS tokenization that aligns the generation order with surface topology; an expansive prediction strategy that maintains coherent, connected surface growth; and a sparse-attention global memory that provides an effectively unbounded receptive field to resolve long-range topological dependencies. This integrated design enables MeshRipple to generate meshes with high surface fidelity and topological completeness, outperforming strong recent baselines.