Fig. 5: Cultivar identification results of different datasets.

a The distribution of F1-scores on the sweet cherry dataset. b Boxplot of the accuracy of cultivar identification using leaves from each period in the soybean dataset independently. The triangle represents the mean accuracy of the corresponding period. The different letters on the top of each box indicate significant differences in accuracies among different growth stages (P < 0.05) using one-way ANOVA