Hyperspectral Image Classification on Botswana 50/50 (test)
96.9Mean Test AccuracyScratch
Evaluation results
| Method | Links | |
|---|---|---|
| ScratchModel Type=Scratch, Backbone=DenseNet121, Rank (R)=2, Number of trainable parameters=456k2026.05 | 96.9 | |
| CPModel Type=CP, Backbone=DenseNet121, Rank (R)=2, Number of trainable parameters=20k2026.05 | 96.6 | |
| TuckerModel Type=Tucker, Backbone=DenseNet121, Rank (R)=2, Number of trainable parameters=20k2026.05 | 96.3 | |
| ReduceModel Type=Reduce, Backbone=DenseNet121, Rank (R)=2, Number of trainable parameters=20k2026.05 | 89.7 |