OPV2V: An Open Benchmark Dataset and Fusion Pipeline for Perception with Vehicle-to-Vehicle Communication

Benchmark Model Rank Results
3d-object-detection-on-opv2vAttentive Fusion (PointPillar backbone)#2AP@0.7@Default: 0.815AP@0.7@CulverCity: 0.735
3d-object-detection-on-opv2vLate Fusion (PointPillar backbone)#5AP@0.7@Default: 0.781AP@0.7@CulverCity: 0.669
3d-object-detection-on-v2xsetAttentiveFusion#6AP0.5 (Perfect): 0.807AP0.7 (Perfect): 0.664AP0.5 (Noisy): 0.709