Visual Localization on Cambridge Landmark (test) benchmark leaderboard · SOTA2 Research

Visual Localization on Cambridge Landmark (test)

17.8Kings Median Translation Error (cm)ACE + UGS-Loc

Evaluation Results

Method	Links
ACE + UGS-Loc2026.03		17.8	0.18	13.8	0.3	4.2	0.16	6.3	0.2	10.5	0.21
RAPref2026.03		18	0.38	22	0.42	8	0.39	16	0.47	16	0.42
DFNet + UGS-Loc2026.03		18.7	0.19	14.5	0.29	3.9	0.15	5.5	0.17	10.7	0.2
GLACE2026.03		20	0.32	20	0.41	5	0.22	9	0.3	14	0.32
ACE + GS-CPR2026.03		20	0.29	21	0.4	5	0.24	13	0.4	15	0.33
DFNet + GS-CPR2026.03		23	0.32	42	0.74	10	0.36	27	0.62	26	0.51
FQN-MN2026.03		28	0.4	54	0.8	13	0.6	58	2	38	1
ACE2026.03		29	0.38	31	0.61	5	0.3	19	0.6	21	0.47
MCLoc2026.03		31	0.42	39	0.73	12	0.45	26	0.8	27	0.6
LENS2026.03		33	0.5	44	0.9	27	1.6	53	1.6	39	1.15
HR-APR2026.03		36	0.58	53	0.89	13	0.51	38	1.16	35	0.78
DFNet + NeFeS302026.03		37	0.64	98	1.61	17	0.6	42	1.38	49	1.06
DFNet + NeFeS502026.03		37	0.54	52	0.88	15	0.53	37	1.14	35	0.77
CrossFire2026.03		47	0.7	43	0.7	20	1.2	39	1.4	37	1
RAP2026.03		58	0.87	81	1.2	33	1.64	60	1.83	58	1.39
DFNet2026.03		73	2.37	200	2.98	67	2.21	137	4.02	119	2.9
MS-Transformer2026.03		85	1.45	175	2.43	88	3.2	166	4.12	129	2.8
PoseNet2026.03		93	2.73	224	7.88	147	6.62	237	5.94	175	5.79