Research Article
Multiscale Dense Cross-Attention Mechanism with Covariance Pooling for Hyperspectral Image Scene Classification
Table 7
Classification results of different methods on the SV data (%).
| Methods | 1% | 5% | 10% | OA | AA | Kappa | OA | AA | Kappa | OA | AA | Kappa |
| AlexNet [23] | 90.88 | 93.72 | 89.77 | 94.05 | 96.64 | 93.42 | 94.26 | 95.67 | 95.29 | ResNet [33] | 87.21 | 92.07 | 85.88 | 91.70 | 95.15 | 90.74 | 93.52 | 96.73 | 93.11 | DenseNet [34] | 85.52 | 90.81 | 84.12 | 91.03 | 94.36 | 90.05 | 93.14 | 96.12 | 92.17 | PRAN [35] | 78.86 | 76.17 | 76.49 | 90.28 | 88.17 | 89.16 | 91.81 | 89.32 | 91.17 | FSSFNet [36] | 91.04 | 94.74 | 90.02 | 93.56 | 96.69 | 92.82 | 95.85 | 98.03 | 95.37 | SAGP [37] | 90.87 | 94.68 | 89.82 | 92.64 | 95.69 | 91.81 | 94.76 | 97.45 | 94.16 | MDCA-CP | 91.45 | 94.84 | 89.98 | 94.51 | 97.11 | 93.24 | 96.15 | 98.11 | 96.21 |
|
|