Hyperspectral Image Classification on CASI University of Houston (test)
86.61Overall AccuracyFreeNet
Evaluation results
| Method | Links | ||||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| FreeNetInput Patch Strategy=Patch-free2020.11 | 86.61 | 80.91 | 84.21 | 98.02 | 91.95 | 100 | 96.5 | 88.53 | 74.83 | 87.72 | 62.25 | 83.4 | 98.84 | 88.42 | 96.76 | 94.29 | 88.44 | 0.8555 | |
| DFFNInput Patch Strategy=Patch-based2020.11 | 84.56 | 77.41 | 81.39 | 94.59 | 87.82 | 96.05 | 96.15 | 80.25 | 77.78 | 84.66 | 64.63 | 88.61 | 98.57 | 83.09 | 99.72 | 81.9 | 86.18 | 0.8328 | |
| Gabor-CNNInput Patch Strategy=Patch-based2020.11 | 84.32 | 82.3 | 84.24 | 95.61 | 92.39 | 99.8 | 96.47 | 84.58 | 73.09 | 78.14 | 58.01 | 73.35 | 86.74 | 91.16 | 100 | 77.73 | 84.17 | 0.8114 | |
| S-CNNInput Patch Strategy=Patch-based2020.11 | 82.34 | 83 | 83.27 | 98.66 | 93.5 | 96.91 | 94.12 | 79.95 | 68.19 | 76.39 | 48.1 | 74.64 | 85.49 | 88.58 | 99.19 | 96.4 | 84.43 | 0.8052 | |
| CMTLlabeled samples per class=52026.01 | 81.38 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 83.18 | 0.7988 | |
| 3D-GANInput Patch Strategy=Patch-based2020.11 | 78.16 | 81.58 | 79.74 | 97.42 | 93.36 | 99.71 | 95.08 | 89.9 | 70.52 | 54.89 | 49.89 | 77.36 | 60.46 | 81.71 | 95.14 | 65.35 | 79.98 | 0.7616 | |
| GMTNlabeled samples per class=52026.01 | 77.2 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 80.11 | 0.7538 | |
| SVMInput Patch Strategy=Patch-based2020.11 | 76.88 | 82.05 | 80.55 | 100 | 92.52 | 98.11 | 95.1 | 75 | 40.17 | 74.88 | 51.64 | 78.37 | 68.4 | 69.47 | 100 | 98.1 | 80.29 | 0.7513 |