Traffic Forecasting on BANGKOK unseen city (test)
0.1423MAERCSNet
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| RCSNetTraining Cities=Berlin, Antwerp, Moscow2026.05 | 0.1423 | 0.5407 | 0.7353 | |
| Multi-task Transfer LearningTraining Cities=Berlin, Antwerp, Moscow2026.05 | 0.1645 | 0.6744 | 0.8212 | |
| UNetTraining Cities=Berlin, Antwerp, Moscow2026.05 | 0.1712 | 0.7362 | 0.858 | |
| SwinUnet3DTraining Cities=Berlin, Antwerp, Moscow2026.05 | 0.1778 | 0.8001 | 0.8945 | |
| U-Net + TDATraining Cities=Berlin, Antwerp, Moscow2026.05 | 0.1823 | 0.8386 | 0.9158 | |
| Traffic Movie SnippetsTraining Cities=Berlin, Antwerp, Moscow2026.05 | 0.1856 | 0.8909 | 0.9439 | |
| 3DResNet + Sparse-UNetTraining Cities=Berlin, Antwerp, Moscow2026.05 | 0.1867 | 0.8962 | 0.9467 | |
| Graph-based U-NetTraining Cities=Berlin, Antwerp, Moscow2026.05 | 0.1934 | 0.967 | 0.9834 | |
| Dual Encoding U-NetTraining Cities=Berlin, Antwerp, Moscow2026.05 | 0.1978 | 1.0088 | 1.0044 |