MeshFlow: Mesh Generation with Equivariant Flow Matching
- Published
- Source
- arXiv
- Paper number
- 483
- Field
- Computer Vision
- arXiv ID
- 2606.23489
Key points
- It directly represents meshes as triangle soup, avoiding the limitations of autoregressive serialization such as slow speed and error accumulation.
- It designs an Equivariant DiT architecture that preserves both face permutation invariance and vertex cyclic invariance.
- It introduces a coupling loss based on nested optimal transport that removes learning signals that violate symmetry.
- It achieves the best 1-NNA on three of the four ShapeNet categories and matches the quality of state-of-the-art autoregressive models.
- It speeds up inference by 18.55x, enabling sub-second mesh generation compared with autoregressive methods.
- A post-processing denoiser reduces self-intersection rate, Ri, by about 56% with negligible additional latency.
Paper links
External research summaries. These are not HDATF publications or measured product results.