A Truncated Transfer Learning Based Recognition Method for COVID-19
This paper uses the residual network ResNet50 as the backbone network and adopts a truncated transfer learning method to address the characteristic of medical diseases that pay more attention to abstract features at the middle and bottom layers.This method preserves and fine tunes some of the lower layers and directly discards other layers.At the same time,a convolutional attention module is added between the truncation module and the fully connected layer to make the model pay more attention to the feature information of the lesion area,achieving fast and accurate recognition of pneu-monia images.Experiments are conducted on the newly collected COVID-Xray15k dataset,and the classification accuracy of the model reached 98.6%.Compared with the existing research,the new model has more accurate and efficient perfor-mance in recognizing COVID-Xray15k images.
万方数据
维普
