Porto
Benchmarks
| Task Name | Dataset Name | SOTA Result | Trend | |
|---|---|---|---|---|
| Grid-based Trajectory Generation | Porto | JSD0.02 | 30 | |
| Road Network-based Trajectory Generation | Porto | JSD7 | 30 | |
| Map extraction | Porto regions (test) | IoU71.7 | 18 | |
| Trajectory Similarity Search | Porto (test) | HR@143.5 | 11 | |
| Estimated Time of Arrival | Porto | MAPE (%)12.42 | 8 | |
| Trajectory Generation | Porto Dataset (test) | Jac52.2 | 8 | |
| Trajectory Generation | Porto | T-Dist0.0002 | 7 | |
| Full-trajectory travel-time regression | Porto (test) | MAE (s)38.41 | 7 | |
| Call-type classification | Porto | Micro-F181.3 | 6 | |
| Hard Braking Event (HBE) Prediction | Porto | MAE0.401 | 6 | |
| AOI Function Prediction | Porto | L1 Distance0.279 | 6 | |
| Road Popularity Prediction | Porto | F1 Score90.8 | 6 | |
| Next-edge prediction | Porto (test) | Cross-Entropy Loss0.88 | 5 | |
| Trajectory Generation | Porto (test) | Density0.0121 | 5 | |
| Road Extraction | Porto (five-fold cross-validation) | Metric- | 0 |
Showing 15 of 15 rows