End-to-End Autonomous Driving Planning on NAVSIM v1 (Navtest)
1NC ScoreHuman
Evaluation Results
| Method | Links | ||||||
|---|---|---|---|---|---|---|---|
| HumanInput=N/A, Backbone=N/A2025.12 | 1 | 1 | 0.875 | 1 | 0.999 | 0.948 | |
| WoTEInput=Camera & LiDAR, Backbone=ResNet-342025.12 | 0.985 | 0.968 | 0.819 | 0.949 | 0.999 | 0.883 | |
| DIVERInput=Camera & LiDAR, Backbone=ResNet-342025.12 | 0.985 | 0.965 | 0.826 | 0.949 | 1 | 0.883 | |
| MindDriveInput=Camera & LiDAR, Backbone=ResNet-342025.12 | 0.984 | 0.97 | 0.828 | 0.951 | 0.999 | 0.889 | |
| Hydra-MDPInput=Camera & LiDAR, Backbone=ResNet-342025.12 | 0.983 | 0.96 | 0.787 | 0.946 | 1 | 0.865 | |
| GoalFlowInput=Camera & LiDAR, Backbone=ResNet-342025.12 | 0.983 | 0.938 | 0.798 | 0.943 | 1 | 0.857 | |
| ARTEMISInput=Camera & LiDAR, Backbone=ResNet-342025.12 | 0.983 | 0.951 | 0.814 | 0.943 | 1 | 0.87 | |
| DiffusionDriveInput=Camera & LiDAR, Backbone=ResNet-342025.12 | 0.982 | 0.962 | 0.822 | 0.947 | 1 | 0.881 | |
| DRAMAInput=Camera & LiDAR, Backbone=ResNet-342025.12 | 0.98 | 0.931 | 0.801 | 0.948 | 1 | 0.855 | |
| PARA-DriveInput=Camera, Backbone=ResNet-342025.12 | 0.979 | 0.924 | 0.793 | 0.93 | 0.998 | 0.84 | |
| VADv2Input=Camera, Backbone=ResNet-342025.12 | 0.979 | 0.917 | 0.776 | 0.929 | 1 | 0.83 | |
| UniADInput=Camera, Backbone=ResNet-342025.12 | 0.978 | 0.919 | 0.788 | 0.929 | 1 | 0.834 | |
| TransFuserInput=Camera & LiDAR, Backbone=ResNet-342025.12 | 0.977 | 0.928 | 0.792 | 0.928 | 1 | 0.84 | |
| Hydra-MDP++Input=Camera & LiDAR, Backbone=ResNet-342025.12 | 0.976 | 0.96 | 0.804 | 0.931 | 1 | 0.866 | |
| LTFInput=Camera, Backbone=ResNet-342025.12 | 0.974 | 0.928 | 0.79 | 0.924 | 1 | 0.838 | |
| LAWInput=Camera, Backbone=ResNet-342025.12 | 0.964 | 0.954 | 0.817 | 0.887 | 0.999 | 0.846 |