3D Object Detection on KITTI (Class & Difficulty APs) benchmark leaderboard · SOTA2 Research

3D Object Detection on KITTI (Class & Difficulty APs)

92.563D AP (Car, Moderate)PDV

Evaluation Results

Method	Links
PDV2026.05		92.56	85.29	83.05	60.8	66.9	55.85	74.23	92.72	69.6
VoxelNextFusion2026.05		86.98	92.78	84.59	—	—	—	—	—	—
SE-SSD2026.05		86.25	90.21	79.22	—	—	—	—	—	—
VFF2026.05		85.51	92.31	82.92	65.11	73.26	60.03	73.12	89.4	69.68
FocalsConv2026.05		85.32	92.26	82.95	—	—	—	—	—	—
Voxel-RCNN2026.05		85.29	92.38	82.86	—	—	—	—	—	—
Ours2026.05		85.21	92.98	83.04	68.42	74.99	63.34	75.41	92.15	73.11
LoGoNet2026.05		85.04	92.04	84.31	63.72	70.2	59.46	75.35	91.74	72.42
PV-RCNN2026.05		84.36	92.1	82.48	56.67	64.26	51.91	71.95	88.88	66.78
OneForAll2026.05		84.2	92.8	82.3	—	—	—	—	—	—
CAT-Det2026.05		81.46	90.12	79.15	66.35	74.08	58.92	72.82	87.64	68.2
MENet2026.05		78.82	89.14	66.33	74.79	59.8	66.27	85.04	62.73	—
EPNet2026.05		78.65	88.76	78.32	59.29	66.74	54.82	65.6	83.88	62.7
SECOND2026.05		78.62	88.61	77.22	52.98	56.55	47.73	67.15	80.58	63.1
PointPillars2026.05		77.28	86.46	74.65	52.29	57.57	47.9	62.28	80.05	59.7