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

Multi-Label Remote Sensing Image Retrieval on DLRSD

77.71mAP (sim) @ 5000Wg-MACL

Evaluation Results

Method	Links
Wg-MACL2025.12		77.71	80.2	29.26	13.19	5.31	77.88	19.13
LReg2025.12		75.23	77.6	29.64	13.6	5.96	74.69	15.57
OML2025.12		73.27	76.08	29.28	13.13	5.07	71.58	18.17
MACL2025.12		72.14	75.17	28.95	14.28	6.85	73.92	16.13
L-W/O-Reg2025.12		70.88	73.68	30.25	13.72	5.58	74.72	15.27
Wg-MulSupCon2025.12		70.32	73.5	30.34	13.69	5.47	74.88	15.17
LProto2025.12		70.03	73.14	30.25	13.71	5.43	74.47	15.34
MulSupCon2025.12		69.71	72.51	29.7	13.27	5.01	74.43	14.67
LBase2025.12		69.14	71.43	30.36	13.47	5.12	74	14.31
LMsc2025.12		69.09	71.54	30.14	13.41	5.27	74.15	14.82
SupCon ALL2025.12		68.72	71.14	30.12	13.45	5.32	73.27	14.3
SupCon ANY2025.12		68.43	69.85	30.13	13.55	5.49	72.42	13.76
Sim-Dis Loss2025.12		67.74	70.01	30.12	13.55	5.31	72.67	13.37