Open-set Object Detection on Cityscapes to BDD100k 5 novel categories benchmark leaderboard · SOTA2 Research

Open-set Object Detection on Cityscapes to BDD100k 5 novel categories

16.89mAPCCKM

Evaluation Results

Method	Links
CCKM2026.04		16.89	12.94	1.024	6,152
SOMA2026.04		16.63	15.59	1.181	7,230
OpenDet2026.04		16.11	14.69	1.25	7,988
DDETR2026.04		15.86	0	1.258	13,044
OW-DETR2026.04		15.81	9.6	1.379	7,738
CCKM2026.04		14.37	1	0.109	626
SOMA2026.04		14.13	2.01	0.197	1,074
OpenDet2026.04		13.75	1.27	0.209	1,244
DDETR2026.04		13.52	0	0.227	2,378
OW-DETR2026.04		13.5	1.3	0.201	1,168
PROSER2026.04		13.37	1.6	0.218	1,466
PROSER2026.04		12.88	15.57	1.286	7,504
CCKM2026.04		12.06	2.42	1.861	5,966
SOMA2026.04		11.83	4.13	2.611	9,968
CCKM2026.04		11.75	2.6	2.67	3,286
SOMA2026.04		11.71	4.53	2.713	8,844
OpenDet2026.04		10.84	3.41	2.861	11,776
DDETR2026.04		10.56	0	3.003	29,812
OW-DETR2026.04		10.52	1.64	2.78	10,088
OpenDet2026.04		10.46	1.73	3.393	15,928
DDETR2026.04		10.18	0	3.219	25,684
PROSER2026.04		9.8	1.77	4.202	28,170
PROSER2026.04		9.4	2.58	3.067	15,962
OW-DETR2026.04		8.34	1.76	2.867	11,034