BEV Object Detection on TJ4DRadSet (test) benchmark leaderboard · SOTA2 Research

BEV Object Detection on TJ4DRadSet (test)

54.07BEV mAPR4Det

Evaluation Results

Method	Links
R4Det2026.03		54.07	72.36	33.48	64.58	45.85
Ours2026.01		49.33	67.73	34.94	61.32	33.4
SGDet3D2026.01		47.16	66.38	29.18	53.72	39.36
SGDet3D2026.03		47.16	66.38	29.18	53.72	39.36
RCDFNet2026.01		45.61	54.42	35.81	59.27	32.93
SFGFusion2026.01		44.66	46.63	30.04	59.55	42.21
CVFusion2026.03		44.07	58.07	31.65	51.29	35.29
HSGFusion2026.01		43.23	—	—	—	—
UniBEVFusion2026.01		42.92	50.43	29.57	56.48	35.22
LXLv22026.01		42.35	—	—	—	—
Chen et al.2026.01		41.75	43.73	34.08	59.49	29.7
MAFF-Net2026.01		41.59	39.05	35.25	56.35	35.73
BEVFusion2026.01		41.23	48.53	32.04	55.4	28.96
DSFusion2026.01		41.21	49.86	34.17	52.49	32.39
LXL2026.01		41.2	—	—	—	—
LXL2026.03		41.2	—	—	—	—
BEVFusion2026.03		41.12	—	—	—	—
SMURF2026.01		40.98	43.13	29.19	58.81	32.8
SMURF2026.03		40.98	43.13	29.19	58.81	32.8
RCFusion2026.01		39.76	40.89	30.95	58.3	28.92
RCFusion2026.03		39.76	40.89	30.95	58.3	28.92
RadarPillarNet2026.01		39.24	45.72	29.19	56.89	25.17
RadarPillarNet2026.03		39.24	45.72	29.19	56.89	25.17
PointAugmenting2026.01		39.04	43.42	29.65	59.21	23.88
FUTR3D2026.01		37.51	—	—	—	—
FUTR3D2026.03		37.51	—	—	—	—
RCBEVDet2026.01		37.46	41.77	28.98	53.03	26.05
PointPillars2026.01		36.23	38.34	32.26	56.11	18.19
PointPillars2026.03		36.23	38.34	32.26	56.11	18.19
MUFASA2026.01		36.19	41.72	27.93	46.42	28.69
CenterPoint2026.01		36.18	33.03	27.87	58.74	25.09
CenterPoint2026.03		36.18	33.03	27.87	58.74	25.09