Multi-label Remote Sensing Image Retrieval on ML-AID benchmark leaderboard · SOTA2 Research

Multi-label Remote Sensing Image Retrieval on ML-AID

95.77mAP (sim) @ 5000MACL

Evaluation Results

Method	Links
MACL2025.12		95.77	95.9	59.89	38.86	14.91	77.12	43.84
Wg-MACL2025.12		94.46	94.43	58.2	37.94	14.54	89.84	39.42
OML2025.12		93.13	93.52	57.9	38.2	15	67.61	36.78
LReg2025.12		91.88	92.56	56.35	36.86	14.73	85.94	32.95
L-W/O-Reg2025.12		91.81	92.76	56.24	36.56	14.15	87.85	36.98
Wg-MulSupCon2025.12		90.11	90.21	56.24	36.81	14.44	84.43	30.6
MulSupCon2025.12		89.62	90.09	55.56	36.63	14.46	84.61	30.84
LMsc2025.12		89.1	89.77	56.33	36.96	14.51	83.83	29.71
LBase2025.12		88.98	89.96	55.47	36.32	14.21	84.39	31.14
LProto2025.12		88.41	89.64	56.18	36.94	14.73	83.98	30.28
Sim-Dis Loss2025.12		88.36	89.04	56.91	37.37	14.76	58.04	29.45
SupCon ANY2025.12		87.2	86.31	56.95	37.75	14.93	82.22	26.9
SupCon ALL2025.12		86.95	89.32	56.36	37.06	14.6	84.33	30.75