Skip to main content
. 2020 Apr 13;15(4):e0230287. doi: 10.1371/journal.pone.0230287

Table 2. The comparison results of CNN model and traditional identification model.

Networks Train_accuracy Train_Loss Test_accuracy Test_Loss
VGG16 0.983 0.085 0.9441 0.156
DenseNets121 0.998 0.009 0.9256 0.271
InceptionV4 1.00 0.00 0.9744 0.201
ResNet10 0.999 0.001 0.9935 0.024