Multi-Label Remote Sensing Image Retrieval on WHDLD benchmark leaderboard · SOTA2 Research

Multi-Label Remote Sensing Image Retrieval on WHDLD

94.41mAP (sim)@5000MACL

Evaluation Results

Method	Links
MACL2025.12		94.41	93.72	82.54	66.22	42.31	79.42	63.24
MACL2025.12		94.14	93.55	83.23	67.45	43.1	78.2	61.9
LReg2025.12		93.88	92.62	82.74	67.22	43.26	76.28	59.63
LMsc2025.12		93.87	92.59	82.21	66.34	41.98	77.2	60.48
MulSupCon2025.12		93.83	92.69	82.63	66.98	42.81	76.72	60.08
MulSupCon2025.12		93.46	92.23	83.15	67.09	42.6	74.53	57.72
LBase2025.12		93.23	92.46	82.81	67.32	43.22	76.59	59.78
SupCon ANY2025.12		92.99	92.22	82.22	66.02	41.53	74.48	57.12
OML2025.12		92.65	91.22	81.43	64.78	40.36	70.8	52.37
SupCon ALL2025.12		90.87	92.77	82.86	66.66	42.26	78.06	61.54
LReg2025.12		90.71	89.68	82.1	65.88	41.49	65.94	47.5
LProto2025.12		90.62	92.47	82.91	66.45	41.75	73.9	56.5
Sim-Dis Loss2025.12		89.77	90.42	82.64	66.52	42.05	69.71	52.35