List your company

Motion Planning on nuScenes

0.14L2 Error (1s)OmniDrive
0.08760.44130.7951.1487Mar 30, 2026Apr 15, 2026May 1, 2026May 18, 2026Jun 3, 2026Jun 19, 2026Jul 6, 2026

Evaluation results

MethodLinks
OmniDriveMethod Category=Autoregressive Methods, VLM=LLava-7B2026.030.14————————0.290.550.33140.2711
OpenDriveVLAMethod Category=Autoregressive Methods, VLM=Qwen2.5-VL-3B2026.030.14————————0.30.550.33270.2210
AutoDrive-P³Method Category=Autoregressive Methods, Variant=Detailed, VLM=Qwen2.5-VL-3B2026.030.15————————0.30.540.33020.156
AutoDrive-P³Method Category=Autoregressive Methods, Variant=Fast, VLM=Qwen2.5-VL-3B2026.030.16————————0.310.560.34040.28
VADMethod Category=Non-Autoregressive Methods2026.030.17————————0.340.60.377100.2414
RDA-DriverMethod Category=Autoregressive Methods, VLM=LLaVa-7B2026.030.17————————0.370.690.4150.2610
DriveVLMMethod Category=Autoregressive Methods, VLM=Qwen2-VL-7B2026.030.18————————0.340.680.410220.4527
GPT-DriverMethod Category=Autoregressive Methods, VLM=GPT-3.52026.030.2————————0.40.70.444120.3617
InsightDriveMethod Category=Non-Autoregressive Methods2026.030.23————————0.410.680.449100.2715
RCT-AD-LBackbone=R101, GPU Type=NVIDIA RTX A6000, Batch Size=12026.070.23————————0.520.890.5413156
AutoVLAMethod Category=Autoregressive Methods, VLM=Qwen2.5-VL-3B2026.030.25————————0.460.730.48770.2613
BridgeADBackbone=R101, GPU Type=NVIDIA RTX A6000, Batch Size=12026.070.28————————0.550.920.5804208
RCT-AD-TBackbone=R50, GPU Type=NVIDIA RTX A6000, Batch Size=12026.070.29————————0.570.930.5914187
SparseDriveBackbone=R50, Official Checkpoint Evaluation=true, GPU Type=NVIDIA RTX A6000, Batch Size=12026.070.3————————0.580.950.61152310
AutoDrive-R²Method Category=Autoregressive Methods, VLM=Qwen2.5-VL-3B2026.030.35————————0.490.620.49————
GenADBackbone=R50, GPU Type=NVIDIA RTX A6000, Batch Size=12026.070.36————————0.831.550.9162310043
VADBackbone=R50, Official Checkpoint Evaluation=true, GPU Type=NVIDIA RTX A6000, Batch Size=12026.070.41————————0.71.050.723194321
UniADMethod Category=Non-Autoregressive Methods2026.030.44————————0.670.960.69480.2312
UniADBackbone=R101, Official Checkpoint Evaluation=true, GPU Type=NVIDIA Tesla A100, Batch Size=12026.070.45————————0.71.040.735177131
Ego-MLPMethod Category=Non-Autoregressive Methods2026.030.46————————0.761.120.7821350.5838
ST-P3Method Category=Non-Autoregressive Methods2026.031.33————————2.112.92.1123621.2771
OpenEMMAMethod Category=Autoregressive Methods, VLM=Qwen2-VL-7B2026.031.45————————3.213.762.81————
AutoVLAParadigm=AR, Backbone=Qwen2.5-3B2025.12—0.130.180.280.20.140.250.530.31———————
DME-DriverParadigm=AR, Backbone=LLaVA-7B2025.12—————0.050.280.550.29———————
DriveVLMParadigm=AR, Backbone=Qwen2-VL-7B2025.12—0.10.220.450.27———————————
Ego-MLP2025.12—0.210.350.580.38———————————
EponaParadigm=AR + Diff., Backbone=DiT-2.5B2025.12—0.050.220.850.96———————————
GPT-DriverParadigm=AR, Backbone=GPT-42025.12—0.040.120.360.170.070.151.10.44———————
InsightDrive2025.12—0.090.10.270.150.080.150.840.36———————
OmniDriveParadigm=AR, Backbone=LLaVA-7B2025.12—0.040.462.320.94———————————
PreWorld2025.12—————0.190.572.651.14———————
ST-P32025.12—0.230.621.270.71———————————
UniAD2025.12—0.040.080.230.120.050.170.710.31———————
VAD-v22025.12—0.070.10.240.14———————————
WAM-FlowParadigm=DFM, Backbone=Janus-1.5B2025.12—0.040.10.230.1200.10.60.23———————