Autonomous Driving on CARLA Town05 (Long) (Metrics: DS, RC, IS)
85.34DSM2DA
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| M2DA2026.05 | 85.34 | — | — | |
| VADv2Modality=C2024.02 | 85.1 | 98.4 | 87 | |
| ReasonNet2026.05 | 79.95 | 89.89 | 89 | |
| InterFuserDVS2026.05 | 77.2 | 100 | 77 | |
| InterFuser2026.05 | 76.18 | 88.23 | 84 | |
| DriveMLMModality=C+L2024.02 | 76.1 | 98.1 | 78 | |
| TCP2026.05 | 75.14 | 85.63 | 87 | |
| DriveAdapter + TCPTeacher=Reinforcement Learning, Student=Frozen Teacher + Adapter, Dual output head=true, Extra data=true2023.08 | 71.9 | 97.3 | 74 | |
| DriveAdapter+TCPModality=C+L2024.02 | 71.9 | 97.3 | 74 | |
| ThinkTwiceEncoder=Geometric Fusion in BEV, Decoder=Look-Predict-Refine, Modality=C4L1, Extra Supervision=Expert+Dep+Seg+Map, using extra data=true2023.05 | 70.9 | 95.5 | 75 | |
| ThinkTwiceTeacher=Reinforcement Learning, Student=Behavior Cloning, Extra data=true2023.08 | 70.9 | 95.5 | 75 | |
| ThinkTwiceModality=C+L2024.02 | 70.9 | 95.5 | 75 | |
| LAWSupervision=Expert, Latent Prediction2024.06 | 70.1 | 97.8 | 72 | |
| InterfuserEncoder=Fusion via Transformer, Decoder=Transformer + GRU, Modality=C3L1, Extra Supervision=Map+Box, using extra data=true2023.05 | 68.3 | 95 | — | |
| InterfuserTeacher=Rule-Based, Student=Behavior Cloning + Rule, Extra data=true2023.08 | 68.3 | 95 | — | |
| InterfuserModality=C2024.02 | 68.3 | 95 | — | |
| InterfuserSupervision=Expert, Map., Det.2024.06 | 68.3 | 95 | — | |
| DriveAdapter + TCPTeacher=Reinforcement Learning, Student=Frozen Teacher + Adapter, Dual output head=true2023.08 | 65.9 | 94.4 | 72 | |
| DriveAdapterSupervision=Expert, Map., Det.2024.06 | 65.9 | 94.4 | 72 | |
| ThinkTwiceEncoder=Geometric Fusion in BEV, Decoder=Look-Predict-Refine, Modality=C4L1, Extra Supervision=Expert+Dep+Seg+Map2023.05 | 65 | 95.5 | 69 | |
| ThinkTwiceTeacher=Reinforcement Learning, Student=Behavior Cloning2023.08 | 65 | 95.5 | 69 | |
| ThinkTwiceSupervision=Expert, Dep., Seg., Det.2024.06 | 65 | 95.5 | 69 | |
| LAV2026.05 | 61.85 | 94.46 | 64 | |
| DriveAdapterTeacher=Reinforcement Learning, Student=Frozen Teacher + Adapter2023.08 | 61.7 | 92.3 | 69 | |
| TransFuser2026.05 | 61.18 | 86.69 | 71 | |
| MILEEncoder=ResNet + Flatten, Decoder=GRU, Modality=C1, Extra Supervision=Map+Box, using extra data=true, no scenarios in evaluation=true2023.05 | 61.1 | 97.4 | 63 | |
| MILETeacher=Reinforcement Learning, Student=Model-Based Imitation Learning, Extra data=true, No scenarios=true2023.08 | 61.1 | 97.4 | 63 | |
| MILEModality=C2024.02 | 61.1 | 97.4 | 63 | |
| MILESupervision=Expert, Map., Det.2024.06 | 61.1 | 97.4 | 63 | |
| TCPEncoder=ResNet + Flatten, Decoder=GRU, Modality=C1, Extra Supervision=Expert2023.05 | 57.2 | 80.4 | 73 | |
| TCPTeacher=Reinforcement Learning, Student=Behavior Cloning2023.08 | 57.2 | 80.4 | 73 | |
| TCPSupervision=Expert2024.06 | 57.2 | 80.4 | 73 | |
| LAVEncoder=PointPaiting, Decoder=Multi-layer GRUS, Modality=C4L1, Extra Supervision=Expert+Seg+Map+Box2023.05 | 46.5 | 69.8 | 73 | |
| LAVTeacher=Imitation Learning, Student=Behavior Cloning2023.08 | 46.5 | 69.8 | 73 | |
| LAVSupervision=Expert, Seg., Map., Det.2024.06 | 46.5 | 69.8 | 73 | |
| RoachEncoder=ResNet + Flatten, Decoder=MLP, Modality=C1, Extra Supervision=Expert2023.05 | 41.6 | 96.4 | 43 | |
| RoachTeacher=Reinforcement Learning, Student=Behavior Cloning + DAgger2023.08 | 41.6 | 96.4 | 43 | |
| RoachModality=C2024.02 | 41.6 | 96.4 | 43 | |
| RoachSupervision=Expert2024.06 | 41.6 | 96.4 | 43 | |
| TransfuserEncoder=Fusion via Transformer, Decoder=GRU, Modality=C3L1, Extra Supervision=Dep+Seg+Map+Box2023.05 | 31 | 47.5 | 77 | |
| TransfuserTeacher=Rule-based, Student=Behavior Cloning2023.08 | 31 | 47.5 | 77 | |
| TransfuserModality=C+L2024.02 | 31 | 47.5 | 77 | |
| TransfuserSupervision=Expert, Dep., Seg., Map., Det.2024.06 | 31 | 47.5 | 77 | |
| VADModality=C2024.02 | 30.3 | 75.2 | — | |
| LBCEncoder=ResNet + Flatten, Decoder=MLP, Modality=C3, Extra Supervision=Expert2023.05 | 12.3 | 31.9 | 66 | |
| LBCTeacher=Imitation Learning, Student=Behavior Cloning + DAgger2023.08 | 12.3 | 31.9 | 66 | |
| LBCModality=C2024.02 | 12.3 | 31.9 | 66 | |
| LBCSupervision=Expert2024.06 | 12.3 | 31.9 | 66 | |
| ST-P3Modality=C2024.02 | 11.5 | 83.2 | — | |
| CILRSEncoder=ResNet + Flatten, Decoder=MLP, Modality=C1, Extra Supervision=None2023.05 | 7.8 | 10.3 | 75 | |
| CILRSTeacher=Rule-Based, Student=Behavior Cloning2023.08 | 7.8 | 10.3 | 75 | |
| CILRSModality=C2024.02 | 7.8 | 10.3 | 75 | |
| CILRSSupervision=Expert2024.06 | 7.8 | 10.3 | 75 |