Research Article
Hyperspectral Image Classification Model Using Squeeze and Excitation Network with Deep Learning
Table 5
Evaluation of accuracy on each class of Pavia Centre dataset with overall accuracy.
| Class | VGG-16 | Inception-v3 | ResNet-50 | Proposed |
| Water | 99.89 | 100 | 100 | 100 | Trees | 94.85 | 95.78 | 95.10 | 95.43 | Asphalt | 93.90 | 96.39 | 96.02 | 95.18 | Self-blocking bricks | 87.56 | 89.08 | 90.17 | 90.38 | Bitumen | 95.44 | 96.71 | 96.50 | 97.84 | Tiles | 96.34 | 98.58 | 98.49 | 98.95 | Shadows | 95.04 | 95.20 | 94.99 | 95.46 | Meadows | 97.63 | 98.05 | 98.57 | 99.65 | Bare soil | 96.50 | 96.89 | 98.17 | 99.68 | Overall accuracy | 96.93 | 97.90 | 97.55 | 98.94 |
|
|