[Retracted] Identification of Navel Orange Diseases and Pests Based on the Fusion of DenseNet and Self-Attention Mechanism

Table 4

Experimental results of different methods on the data set.

Models	Top-1 accuracy (%)	Top-5 accuracy (%)	Params (M)
AlexNet	87.536	99.396	57.033
VGG19	90.676	99.710	39.599
ResNet18	92.705	99.565	11.180
DenseNet-121	92.729	99.782	6.961
DCPSNET	94.275	99.997	9.995