Improved Object Proposals with Geometrical Features for Autonomous Driving

Table 3

Average Precision (AP) (in %) on the validation set of the KITTI object detection benchmark with 1000 proposals, while, for EB and SS, the number of proposals is 2000.

Metric	Method	Cars			Cyclist			Pedestrian
		Easy	Moderate	Hard	Easy	Moderate	Hard	Easy	Moderate	Hard
AP	SS [16]	75.91	60.00	50.98	56.23	39.16	38.83	54.06	47.55	40.56
	EB [16]	86.81	70.47	61.16	55.01	37.87	35.80	57.79	49.99	42.19
	3DOP	94.47	87.09	78.72	84.65	57.38	55.63	72.47	65	57.24
	Our-SS	95.36	87.84	78.57	84.71	57.74	55.8	74.23	66.54	57.9
	Our-EB	88.92	87.40	78.43	83.38	57.72	55.69	74.39	66.73	58.17