3D Lane Detection on OpenLane 1.5m threshold (val) benchmark leaderboard · SOTA2 Research

3D Lane Detection on OpenLane 1.5m threshold (val)

60.5F-ScoreM²-3DLaneNet

Evaluation Results

Method	Links
M²-3DLaneNet2022.09		60.5	0.396	0.405	0.311	0.326
M²-3DLaneNet2022.09		60.5	0.396	0.405	0.311	0.326
M² C-Stream2022.09		59.2	0.402	0.42	0.313	0.335
Persformer + LiDAR+2022.09		56.9	0.453	0.467	0.354	0.357
Persformer + LiDAR2022.09		56.9	0.452	0.467	0.354	0.357
M²-3DLaneNet2022.09		55.5	0.431	0.487	0.327	0.401
BEVFusion2022.09		54.7	0.455	0.431	0.319	0.334
PointPainting (RGB)2022.09		54.3	0.448	0.443	0.351	0.366
LiDAR+2022.09		53.8	0.462	0.435	0.332	0.343
Persformer2022.09		53.3	0.478	0.563	0.363	0.408
Persformer2022.09		53.3	0.478	0.563	0.363	0.408
Persformer + LiDAR2022.09		53	0.458	0.519	0.357	0.416
SFD2022.09		52.5	0.452	0.441	0.331	0.347
PointPainting (SEG)2022.09		51.5	0.376	0.392	0.303	0.33
M² C-Stream2022.09		50.9	0.433	0.536	0.333	0.452
Persformer2022.09		50.5	0.485	0.553	0.364	0.431
LiDAR†-2022.09		49.5	0.49	0.459	0.36	0.357
3DLaneNet2022.09		44.1	0.479	0.572	0.367	0.443
GenLaneNet2022.09		32.3	0.591	0.684	0.411	0.521