Table 8.

Comparison of trash sorting models.

Datasets	Models	Accuracy (%)	Recall (%)	Kappa (%)	Precision (%)	F1-Score (%)
TrashNet	AlexNet [6]	90.08 ± 0.38	88.65 ± 0.40	87.52 ± 0.28	87.37 ± 0.45	87.81 ± 0.52
	DenseNet169 [7]	95.75 ± 0.29	95.39 ± 0.37	94.77 ± 0.35	95.04 ± 0.33	94.62 ± 0.33
	EfficientNet-B2 [10]	94.13 ± 0.20	93.56 ± 0.36	92.78 ± 0.22	92.88 ± 0.30	93.17 ± 0.26
	Optimized DenseNet121 [8]	94.94 ± 0.22	94.25 ± 0.86	93.77 ± 0.28	93.95 ± 0.50	93.77 ± 0.42
	AM-b Xception [9]	94.57 ± 0.35	94.07 ± 0.34	93.32 ± 0.42	93.63 ± 0.60	93.34 ± 0.85
	ResNeXt50+IPAM	96.05 ± 0.14	95.50 ± 0.43	94.94 ± 0.26	94.50 ± 0.17	95.16 ± 0.17
TrashIVL-5	AlexNet [6]	88.86 ± 0.33	87.16 ± 0.48	87.15 ± 0.39	87.21 ± 0.37	85.56 ± 0.43
	DenseNet169 [7]	96.15 ± 0.18	95.63 ± 0.11	95.76 ± 0.12	95.82 ± 0.25	95.26 ± 0.23
	EfficientNet-B2 [10]	96.22 ± 0.14	95.65 ± 0.06	95.74 ± 0.08	95.85 ± 0.98	95.10 ± 0.18
	Optimized DenseNet121 [8]	93.46 ± 0.91	93.93 ± 1.92	94.42 ± 1.38	94.46 ± 1.11	93.49 ± 1.13
	AM-b Xception [9]	96.10 ± 0.18	95.31 ± 0.19	95.47 ± 0.25	95.66 ± 0.31	94.94 ± 0.23
	ResNeXt50+IPAM	97.42 ± 0.14	96.88 ± 0.09	96.36 ± 0.18	97.12 ± 0.16	96.99 ± 0.11
TrashIVL-12	AlexNet [6]	84.16 ± 0.23	82.11 ± 0.44	82.06 ± 0.37	82.46 ± 0.52	82.14 ± 0.26
	DenseNet169 [7]	91.93 ± 0.26	91.52 ± 0.30	91.19 ± 0.18	90.05 ± 0.28	90.16 ± 0.20
	EfficientNet-B2 [10]	91.91 ± 0.22	91.57 ± 0.25	90.14 ± 0.25	91.04 ± 0.20	90.67 ± 0.26
	Optimized DenseNet121 [8]	92.33 ± 0.35	92.25 ± 0.26	92.27 ± 0.24	91.89 ± 0.23	92.04 ± 0.13
	AM-b Xception [9]	93.98 ± 0.23	92.78 ± 0.38	93.21 ± 0.26	92.91 ± 0.28	93.17 ± 0.18
	ResNeXt50+IPAM	94.08 ± 0.11	93.69 ± 0.08	93.80 ± 0.14	94.01 ± 0.22	93.32 ± 0.30