Table 10.

Performance comparison with different patch sizes.

Patch size	Training loss	Training Acc	Validation loss	Validation Acc
Apple dataset41
2	0.3741	0.9912	0.3302	0.9874
4	0.3828	0.9914	0.3722	0.9914
8	0.3728	0.9972	0.3732	0.9923
16	0.3875	0.9905	0.3826	0.9841
Rice dataset34
2	0.3354	0.9942	0.3671	0.9901
4	0.3249	1.0000	0.3521	0.9942
8	0.3228	1.0000	0.3528	0.9970
16	0.3485	0.9912	0.3662	0.9897
Ibean dataset36
2	0.3334	0.9843	0.4264	0.9579
4	0.3157	0.9904	0.4178	0.9617
8	0.3028	0.9965	0.4028	0.9702
16	0.3394	0.9808	0.4349	0.9496
Cassava dataset42
2	0.4417	0.8846	0.6579	0.7319
4	0.4259	0.9155	0.6297	0.7552
8	0.4228	0.9247	0.6328	0.7651
16	0.4491	0.8808	0.6592	0.7296
PlantVillage dataset14
2	0.1507	0.9904	0.2142	0.9902
4	0.1358	0.9947	0.1857	0.9926
8	0.1347	0.9973	0.1836	0.9941
16	0.2268	0.9923	0.1926	0.9917