DoMINO
A point-cloud neural operator for large engineering simulations
Paper title DoMINO: A Decomposable Multi-scale Iterative Neural Operator for Modeling Large Scale Engineering Simulations
Paper arXiv 2501.13350PDF
Encodes the geometry's point cloud into a global grid with multi-scale ball queries, then, at each point where the field is wanted, pulls a local geometry encoding and a small stencil of neighbouring points, much like a finite-volume stencil, and aggregates them to predict surface and volume fields. Points can be mesh nodes or any sampled cloud.
- Data
- DrivAerML
Share on LinkedIn, X or Bluesky, or or .
Know related work that belongs here? Suggest a link
Something wrong on this page? Tell the editor