MeshRipple: Structured Autoregressive Generation of Artist-Meshes
This addresses a critical limitation in 3D mesh generation for applications like digital art and gaming, though it appears incremental as it builds on existing autoregressive approaches.
The paper tackles the problem of autoregressive mesh generation where existing methods break long-range geometric dependencies, causing holes and fragmented components. MeshRipple introduces a frontier-aware tokenization and sparse-attention memory to generate meshes with high surface fidelity and topological completeness, outperforming recent baselines.
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.