Skip to main content

Table 6 Experimental results of different methods

From: Automatic recognition of micronucleus by combining attention mechanism and AlexNet

Method AP F1 AUC
MobileNet [24] 0.589 0.504 0.931
VGG-16 [19] 0.868 0.803 0.989
VGG-Att 0.948 0.786 0.994
GoogLeNet [17] 0.871 0.780 0.988
GoogLeNet-Att 0.875 0.810 0.988
ResNet [18] 0.912 0.804 0.989
ResNet-Att 0.920 0.877 0.993
AlexNet [16] 0.824 0.749 0.984
Alex-light 0.875 0.800 0.990
MSA-Net [25] 0.919 0.808 0.989
Alex-CA 0.883 0.809 0.991
Our method-CE 0.818 0.548 0.985
Our method 0.932 0.811 0.995
  1. The best results in this table are labeled in bold