Table 3 Class-specific accuracies (%) on Indian Pines dataset.
Class | 3D-CNN-LR33 | RNN-GRU-PReTanh34 | Feature-ensemble ND-SVM25 | CNN-MRF35 | HSINet36 | \(\hbox {U}_{\text {Hfe}}\) \(\hbox {SRVAE}_{11}\)37 | 2D-CNN | \(\hbox {SSFNet}_{\text{2D}}\) |
|---|---|---|---|---|---|---|---|---|
1 | 100 | 70.6 | 99.9 ± 0.1 | 86.5 | 100 | 89.6 | 90.7 ± 7.5 | 95.3 ± 4.3 |
2 | 96.3 ± 1.1 | 70.3 | 66.4 ± 1.4 | 91.5 | 66.9 | 89.4 | 97.6 ± 1.3 | 98.4 ± 1.0 |
3 | 99.5 ± 0.7 | 81.5 | 82.8 ± 1.0 | 96.4 | 62.4 | 85.1 | 97.8 ± 1.6 | 98.7 ± 1.2 |
4 | 100 | 90.2 | 89.9 ± 1.2 | 96.2 | 100 | 82.0 | 97.8 ± 2.4 | 99.0 ± 1.8 |
5 | 99.9 ± 0.2 | 92.0 | 94.6 ± 0.6 | 99.5 | 83.2 | 92.6 | 96.4 ± 2.5 | 97.8 ± 2.4 |
6 | 99.8 ± 0.3 | 96.1 | 99.3 ± 0.1 | 99.8 | 98.0 | 96.7 | 98.6 ± 1.3 | 99.3 ± 0.8 |
7 | 100 | 84.8 | 99.9 ± 0.1 | 78.0 | 100 | 34.8 | 84.6 ± 17.6 | 93.0 ± 9.6 |
8 | 100 | 59.6 | 99.6 ± 0.1 | 98.8 | 99.7 | 98.6 | 99.9 ± 0.3 | 99.9 ± 0.1 |
9 | 100 | 86.2 | 99.9 ± 0.1 | 100 | 100 | 93.8 | 84.9 ± 17.5 | 93.9 ± 9.8 |
10 | 98.7 ± 1.0 | 99.4 | 92.2 ± 0.7 | 94.3 | 77.5 | 89.9 | 97.1 ± 1.9 | 98.4 ± 1.4 |
11 | 95.5 ± 1.2 | 85.0 | 77.7 ± 1.0 | 96.5 | 78.4 | 93.2 | 99.0 ± 0.6 | 99.4 ± 0.5 |
12 | 99.5 ± 0.4 | 77.6 | 83.2 ± 1.2 | 91.9 | 75.0 | 85.5 | 96.8 ± 2.0 | 97.6 ± 2.7 |
13 | 100 | 95.6 | 99.8 ± 0.1 | 98.9 | 99.5 | 99.0 | 98.7 ± 2.2 | 99.1 ± 1.6 |
14 | 99.6 ± 0.6 | 84.6 | 95.7 ± 0.2 | 98.4 | 96.5 | 96.7 | 99.7 ± 0.5 | 99.9 ± 0.2 |
15 | 99.5 ± 1.3 | 90.9 | 86.2 ± 1.1 | 91.5 | 69.1 | 80.1 | 98.9 ± 1.4 | 99.2 ± 1.3 |
16 | 99.3 ± 1.08 | 100 | 99.9 ± 1.0 | 97.9 | 100 | 92.9 | 87.8 ± 7.5 | 95.3 ± 5.9 |
OA | 97.6 ± 0.4 | 88.6 | – | 96.1 | 83.0 ± 0.2 | 91.4 | 98.1 ± 0.4 | 98.9 ± 0.4 |
AA | 99.2 ± 0.2 | 85.3 | 91.7 ± 0.1 | 94.8 | 87.9 ± 0.2 | 87.5 | 95.4 ± 3.1 | 97.8 ± 2.8 |
\(\kappa \times 100\) | 97.0 ± 0.5 | 73.7 | – | 95.8 | 81.9 ± 0.2 | 90.2 | – | – |
