CLASSIFICATION OF PNEUMONIA BY MODIFIED DEEPLY SUPERVISED RESNET AND SENET USING CHEST X-RAY IMAGES

Abstract

In Deep Learning, a Convolutional Neural Network (CNN) extracts the features from the visual imagery. These features can be used for various complex tasks such as image classification and segmentation and detection of different objects. The convolutional layers are stacked over each other to form the state-of-the-art models. A modified SENet architecture is introduced in this study to classify pneumonia from chest x-ray images. Six ResNet blocks are connected back to back. The output from the sixth ResNet and the side outputs from the last three ResNets are fused together. This output is fed as input to the SENet block. The validation accuracy of this fusion architecture is 91.84% on chest x-ray images.