Table 3 Impact of attention mechanism over the Flevoland dataset.

From: PolSAR image classification using shallow to deep feature fusion network with complex valued attention

Attention Location

OA (%)

AA (%)

k \(\times\) 100

Without attention

95.14 ± 0.26

94.27 ± 0.37

94.69 ± 0.28

Before fusion

95.84 ± 0.21

94.86 ± 0.68

94.91 ± 0.32

After fusion

96.01 ± 0.04

95.17 ± 0.62

95.64 ± 0.44

Back to article page