Physical Quantity Prediction on real-world geometry dataset
0.3778Drag CoefficientSuperMeshNet
Evaluation results
| Method | Links | ||||
|---|---|---|---|---|---|
| SuperMeshNetNh (number of HR data samples)=40, N (number of LR data samples)=200, Backbone (MPNN)=MGN2026.05 | 0.3778 | 1.4 | 0.0433 | 17.7 | |
| Ground truth HR dataBackbone (MPNN)=MGN2026.05 | 0.3724 | — | 0.0368 | — | |
| Fully supervisedNh (number of HR data samples)=200, N (number of LR data samples)=200, Backbone (MPNN)=MGN2026.05 | 0.3653 | 1.9 | 0.038 | 3.3 |