Table 10 Class-wise comparative performance of Swin-RGB and Swin-NIR models.
From: TomatoRipen-MMT: transformer-based RGB and NIR spectral fusion for tomato maturity grading
Class | Swin-RGB Accuracy (%) | Swin-RGB F1 | Swin-RGB IoU | Swin-NIR Accuracy (%) | Swin-NIR F1 | Swin-NIR IoU |
|---|---|---|---|---|---|---|
Green | 79.4 | 0.78 | 0.7 | 88.1 | 0.84 | 0.67 |
Half-Ripe | 81.6 | 0.8 | 0.72 | 86.4 | 0.83 | 0.66 |
Fully Ripe | 85.2 | 0.83 | 0.75 | 89.2 | 0.86 | 0.69 |
