Research Article
Multiscale Dense Cross-Attention Mechanism with Covariance Pooling for Hyperspectral Image Scene Classification
Table 5
Classification results of different methods on the IP data (%).
| Methods | 5% | 10% | 15% | OA | AA | Kappa | OA | AA | Kappa | OA | AA | Kappa |
| AlexNet [23] | 68.69 | 56.29 | 64.05 | 74.32 | 65.17 | 70.94 | 81.76 | 79.44 | 79.70 | ResNet [33] | 70.67 | 69.69 | 66.04 | 78.40 | 79.22 | 75.10 | 83.33 | 80.76 | 80.95 | DenseNet [34] | 71.23 | 67.55 | 66.66 | 78.37 | 75.29 | 75.87 | 84.66 | 81.37 | 82.58 | PRAN [35] | 72.45 | 73.37 | 69.48 | 77.66 | 73.28 | 74.47 | 82.72 | 76.51 | 80.28 | FSSFNet [36] | 73.75 | 67.95 | 69.98 | 78.66 | 71.26 | 75.47 | 82.61 | 74.48 | 80.05 | SAGP [37] | 73.49 | 76.58 | 73.61 | 78.36 | 80.89 | 76.72 | 81.59 | 86.50 | 82.89 | MDCA-CP | 76.85 | 79.11 | 73.72 | 82.56 | 81.87 | 80.03 | 87.32 | 87.65 | 85.55 |
|
|