MANet: End-to-End Learning for Point Cloud Based on Robust Pointpillar and Multiattention

Table 2

Results on the KITTI test BEV detection benchmark.

Methods	3D mAP	Car			Pedestrian			Cyclist
		Easy	Mod	Hard	Easy	Mod	Hard	Easy	Mod	Hard
VoxelNet [3]	58.25	89.25	79.26	77.39	46.13	40.74	38.11	66.70	54.76	50.55
Pp [6]	66.19	88.35	86.10	79.83	58.6	50.23	47.19	79.14	62.25	56.00
F-P [9]	65.39	88.70	84.00	75.33	58	50.22	47.20	75.38	61.96	54.98
PIXOR [12]	N/A	89.38	87.30	77.97
MV3D [38]	N/A	86.02	76.90	68.49
SECOND [39]	60.56	88.07	79.37	77.95	55.1	46.27	44.76	73.67	56.04	48.78
MANet	69.91	89.21	86.36	83.10	61.4	55.01	51.23	82.81	68.36	63.76