End-to-end Motion Planning on Navsim (test)
98.3NCHydra-MDP-V8192-W-EP
Evaluation results
| Method | Links | ||||||
|---|---|---|---|---|---|---|---|
| Hydra-MDP-V8192-W-EPInputs=LiDAR & Camera, Vocabulary Size=8192, Weighted confidence=True, Ego Progress training=True2024.06 | 98.3 | 96 | 78.7 | 94.6 | 100 | 86.5 | |
| Hydra-MDP-V8192-WInputs=LiDAR & Camera, Vocabulary Size=8192, Weighted confidence=True2024.06 | 98.1 | 96.1 | 77.8 | 93.9 | 100 | 85.7 | |
| Hydra-MDP-V8192Inputs=LiDAR & Camera, Vocabulary Size=81922024.06 | 97.9 | 91.7 | 77.6 | 92.9 | 100 | 83 | |
| Hydra-MDP-V4096Inputs=LiDAR & Camera, Vocabulary Size=40962024.06 | 97.7 | 91.5 | 77.5 | 92.7 | 100 | 82.6 | |
| Hydra-MDP-V8192-PDMInputs=LiDAR & Camera, Vocabulary Size=8192, Learning Target=Overall PDM score2024.06 | 97.5 | 88.9 | 74.8 | 92.5 | 100 | 80.2 | |
| Vadv2-V8192Inputs=LiDAR & Camera, Vocabulary Size=8192, Distance-based imitation loss=True2024.06 | 97.2 | 89.1 | 76 | 91.6 | 100 | 80.9 | |
| Vadv2-V4096Inputs=LiDAR & Camera, Vocabulary Size=4096, Distance-based imitation loss=True2024.06 | 97.1 | 88.8 | 74.9 | 91.4 | 100 | 79.7 | |
| Vadv2-V4096-PPInputs=LiDAR & Camera, Vocabulary Size=4096, Distance-based imitation loss=True, Post-processing=Transfuser perception2024.06 | 97 | 89.1 | 75 | 91.2 | 100 | 79.9 | |
| TransfuserInputs=LiDAR & Camera2024.06 | 96.5 | 87.9 | 73.9 | 90.2 | 100 | 78 | |
| PDM-ClosedInputs=Perception GT2024.06 | 94.6 | 99.8 | 89.9 | 86.9 | 99.9 | 89.1 |