Table 3 ArsenicNetPlus method versus other methods on cassava dataset using 7-fold cross-validation
Method | Accuracy (%) | Recall (%) | Precision (%) | Loss | F1-score (%) |
|---|---|---|---|---|---|
ArsenicNetPlus | \(95.93\) | \(95.91\) | \(95.98\) | 1.244 | \(95.94\) |
V2-Resnet-101 | \(86.90\) | \(86.80\) | \(87.02\) | 0.779 | \(86.92\) |
EfficientNet-B5 | \(92.43\) | \(92.37\) | \(92.46\) | 1.469 | \(92.42\) |
AlexNet | \(62.46\) | \(61.98\) | \(62.94\) | 2.718 | \(62.46\) |
RepVGG-B3g4 | \(93.08\) | \(93.07\) | \(93.14\) | 0.347 | \(93.11\) |