Semantic Understandings for Aerial Images via Multigrained Feature Grouping

Table 3

Performance comparison with state-of-the-art methods on the DFC15 multilabel dataset (%).

Model	EP	ER	EF1	EF2	LP	LR	LF1	LF2
ReNet-50 [40]	84.89	75.64	80	77.33	81.5	59.99	69.11	63.33
ResNet-RBFNN [41]	82.64	78.76	80.65	79.51	72.01	69.85	70.91	70.27
CA-ResNet-LSTM [15]	85.66	75.84	80.45	77.62	83.83	60.05	69.97	63.66
CA-ResNet-BiLSTM [15]	91.93	79.12	85.05	81.39	94.35	62.35	75.08	66.89
ML_GCN [34]	93.52	93.76	93.64	93.71	91.15	90.02	90.58	90.24
MSGM	94.61	92.71	93.65	93.08	91.42	90.70	91.06	90.84