Table 2 The objective indicators on IP dataset (using 3% training, 3% verification, 94% testing).

From: Asymmetric coordinate attention spectral-spatial feature fusion network for hyperspectral image classification

Class

Training

Test

ContextNet

SSRN

FDSSC

DBDA

PyResNet

DBMA

A2S2KResNet

Ours

1

3

43

48.57 ± 0.107

94.20 ± 0.08

95.49 ± 0.03

94.26 ± 0.04

37.61 ± 0.15

97.83 ± 0.03

86.35 ± 0.07

94.04 ± 0.03

2

42

1337

76.51 ± 0.04

91.17 ± 0.02

97.78 ± 0.01

94.91 ± 0.04

40.70 ± 0.10

95.12 ± 0.02

92.94 ± 0.03

95.61 ± 0.02

3

24

777

53.93 ± 0.11

90.05 ± 0.09

95.65 ± 0.02

92.00 ± 0.04

38.94 ± 0.03

97.31 ± 0.02

90.78 ± 0.03

95.65 ± 0.02

4

7

224

44.46 ± 0.30

91.81 ± 0.08

93.67 ± 0.04

89.22 ± 0.10

28.95 ± 0.18

90.74 ± 0.05

90.52 ± 0.09

94.83 ± 0.03

5

14

455

70.28 ± 0.12

96.96 ± 0.01

94.03 ± 0.08

96.93 ± 0.01

67.95 ± 0.10

96.50 ± 0.04

99.08 ± 0.01

99.15 ± 0.08

6

21

689

91.90 ± 0.03

98.13 ± 0.00

96.48 ± 0.01

97.49 ± 0.01

63.44 ± 0.11

97.26 ± 0.02

95.73 ± 0.03

97.06 ± 0.02

7

3

25

26.51 ± 0.05

66.67 ± 0.47

66.35 ± 0.09

70.90 ± 0.01

52.56 ± 0.41

86.48 ± 0.05

86.04 ± 0.20

65.64 ± 0.11

8

14

447

85.27 ± 0.13

96.95 ± 0.01

99.35 ± 0.01

100.0 ± 0.21

61.26 ± 0.43

100.0 ± 0.00

97.81 ± 0.01

100.0 ± 0.00

9

3

16

13.70 ± 0.08

23.81 ± 0.34

54.89 ± 0.15

69.21 ± 0.17

14.44 ± 0.14

62.89 ± 0.23

58.06 ± 0.13

66.18 ± 0.01

10

29

918

81.39 \(\pm\) 0.02

73.56 \(\pm\) 0.06

88.82 \(\pm\) 0.08

93.13 \(\pm\) 0.03

48.30 \(\pm\) 0.18

92.37 \(\pm\) 0.05

86.54 \(\pm\) 0.03

92.09 \(\pm\) 0.03

11

73

2313

79.20 \(\pm\) 0.03

95.35 \(\pm\) 0.03

97.69 \(\pm\) 0.01

96.03 \(\pm\) 0.00

53.19 \(\pm\) 0.12

92.37 \(\pm\) 0.06

91.19 \(\pm\) 0.02

96.75 \(\pm\) 0.01

12

17

563

51.12 \(\pm\) 0.07

89.64 \(\pm\) 0.03

98.12 \(\pm\) 0.01

96.63 \(\pm\) 0.01

52.86 \(\pm\) 0.05

94.43 \(\pm\) 0.03

94.30 \(\pm\) 0.04

95.57 \(\pm\) 0.02

13

6

193

68.89 \(\pm\) 0.06

94.43 \(\pm\) 0.04

92.94 \(\pm\) 0.02

98.90 \(\pm\) 0.01

44.91 \(\pm\) 0.32

99.66 \(\pm\) 0.00

97.01 \(\pm\) 0.01

97.77 \(\pm\) 0.01

14

37

1184

90.52 \(\pm\) 0.02

93.78 \(\pm\) 0.02

96.94 \(\pm\) 0.03

96.24 \(\pm\) 0.02

78.18 \(\pm\) 0.09

96.81 \(\pm\) 0.02

96.47 \(\pm\) 0.02

97.72 \(\pm\) 0.02

15

11

367

53.88 \(\pm\) 0.13

87.03 \(\pm\) 0.13

96.94 \(\pm\) 0.01

96.01 \(\pm\) 0.02

45.10 \(\pm\) 0.23

94.88 \(\pm\) 0.01

86.18 \(\pm\) 0.09

95.21 \(\pm\) 0.01

16

3

84

38.73 \(\pm\) 0.20

66.67 \(\pm\) 0.47

81.45 \(\pm\) 0.20

97.86 \(\pm\) 0.01

45.45 \(\pm\) 0.41

94.21 \(\pm\) 0.07

94.69 \(\pm\) 0.02

93.13 \(\pm\) 0.06

OA (%)

307

9635

76.08 \(\pm\) 0.03

90.70 \(\pm\) 0.01

95.43 \(\pm\) 0.01

95.28 \(\pm\) 0.01

52.80 \(\pm\) 0.11

94.48 \(\pm\) 0.02

92.34 \(\pm\) 0.00

96.02 \(\pm\) 0.00

AA (%)

  

60.93 \(\pm\) 0.06

84.38 \(\pm\) 0.08

90.41 \(\pm\) 0.02

92.48 \(\pm\) 0.02

48.37 \(\pm\) 0.17

93.05 \(\pm\) 0.01

90.23 \(\pm\) 0.03

92.28 \(\pm\) 0.01

Kappa

  

0.7241 \(\pm\) 0.04

0.8941 \(\pm\) 0.02

0.9479 \(\pm\) 0.01

0.9461 \(\pm\) 0.01

0.4427 \(\pm\) 0.14

0.9368 \(\pm\) 0.02

0.9125 \(\pm\) 0.00

0.9546 \(\pm\) 0.01

Training time

  

166.24

236.73

578.67

338.06

507.81

252.64

233.65

99.40

Test time

  

49.29

37.21

23.93

26.77

103.94

22.34

24.25

20.45

Back to article page