Journal of Big Data

Table 10 Experimental results of different models

From: Accurate identification of cashmere and wool fibers based on enhanced ShuffleNetV2 and transfer learning

Models	EfficientNetB0	MobileNetV2	Wide-ResNet50	ShuffleNetV2	CWCNet
Accyracy(%)	94.792	95.312	85.938	95.312	98.438
Precision(%)	94.821	95.658	85.865	95.153	95.000
Recall(%)	94.523	94.823	86.640	95.288	99.090
Params	4.01M	2.23M	66.84M	2.51M	2.51M
FLOPs	0.02G	0.32G	11.44G	0.15G	0.15G
Train time(s)	914.6	614.9	1280.3	539.2	624.6

Back to article page