Table 3.

Statistical measures for dirt classification.

Model	Class	Measurements
		Precision	Recall	F1	Accuracy %
VGG-16	Ash	0.9803	0.9970	0.9886	99.70
	Food	0.9918	0.9675	0.9795	96.65
	Fibre	0.9559	0.9644	0.9601	96.44
	Paper	0.8621	0.9336	0.8964	89.64
	Paint	0.9689	0.9291	0.9485	92.91
	Soil	0.9993	0.9970	0.9981	99.70
	Sand	0.9878	0.9943	0.9910	99.43
	Hair	0.9897	0.9998	0.9948	99.99
	No-dirt	0.9971	0.9941	0.9956	99.41
VGG-11	Ash	0.9900	0.9960	0.9930	99.60
	Food	0.9709	0.9787	0.9748	97.87
	Fibre	0.9847	0.9555	0.9699	95.55
	Paper	0.8482	0.9458	0.8944	94.58
	Paint	0.9914	0.9110	0.9495	91.10
	Soil	0.9983	0.9985	0.9982	99.85
	Sand	0.9906	0.9820	0.9863	98.20
	Hair	0.9622	0.9977	0.9796	99.77
	No-dirt	0.9798	0.9985	0.9891	99.85
MobileNet V2	Ash	0.9950	0.9920	0.9935	99.20
	Food	0.9907	0.9638	0.9770	96.38
	Fibre	0.9700	0.9585	0.9642	95.85
	Paper	0.9108	0.9038	0.9073	90.38
	Paint	0.9707	0.9516	0.9610	95.16
	Soil	0.9963	0.9993	0.9978	99.93
	Sand	0.9843	0.9970	0.9906	99.70
	Hair	0.9589	0.9954	0.9768	99.54
	No-dirt	0.9812	0.9956	0.9883	99.56
ResNet50	Ash	0.9930	0.9880	0.9905	99.30
	Food	0.9956	0.9606	0.9778	96.06
	Fibre	0.9761	0.9703	0.9732	97.03
	Paper	0.9454	0.9392	0.9423	93.92
	Paint	0.9660	0.9758	0.9709	97.58
	Soil	0.9991	0.9985	0.9993	99.85
	Sand	0.9825	0.9973	0.9899	99.73
	Hair	0.9852	0.9965	0.9908	99.65
	No-dirt	0.9963	0.9985	0.9974	99.85
ResNet101	Ash	0.9990	0.9970	0.9980	99.70
	Food	0.9919	0.9755	0.9836	97.55
	Fibre	0.9910	0.9822	0.9866	98.22
	Paper	0.9016	0.9624	0.9310	96.24
	Paint	0.9889	0.9499	0.9690	94.99
	Soil	0.9997	0.9985	0.9993	99.85
	Sand	0.9899	0.9970	0.9934	99.70
	Hair	0.9730	0.9965	0.9846	99.65
	No-dirt	0.9971	0.9985	0.9978	99.85
Darknet53	Ash	0.9979	0.9699	0.9837	96.99
	Food	0.9382	0.9307	0.9344	93.07
	Fibre	0.9939	0.9614	0.9774	93.07
	Paper	0.9001	0.6681	0.7570	66.81
	Paint	0.9861	0.9583	0.9720	95.83
	Soil	0.9901	0.9993	0.9996	99.93
	Sand	0.9568	0.9111	0.9334	91.11
	Hair	0.6506	0.9965	0.7872	99.65
	No-dirt	0.9728	0.9985	0.9855	99.65