Table 5 Impact of different modules on network OA value (%).

From: CMTNet: a hybrid CNN-transformer network for UAV-based hyperspectral crop classification in precision agriculture

Case

CNN Branch

Conv3D

Conv2D

MOCM

LongKou

HanChuan

HongHu

1

\(\times\)

\(\times\)

\(\times\)

\(\times\)

94.97

92.86

91.53

2

\(\checkmark\)

\(\times\)

\(\times\)

\(\times\)

96.25

95.21

93.89

3

\(\checkmark\)

\(\checkmark\)

\(\times\)

\(\times\)

98.77

95.76

95.34

4

\(\checkmark\)

\(\checkmark\)

\(\checkmark\)

\(\times\)

99.12

96.05

97.29

5

\(\checkmark\)

\(\checkmark\)

\(\checkmark\)

\(\checkmark\)

99.58

97.29

98.31