Deep learning in Malaria and Covid19 detection

Abstract

Recent days deep learning is a promising field of research which has been used in biological applications. It is deep learning which is used in biological image detection along with the transformation of analysis data and also in interpreting those data. The revolution of DL enables the researchers and scientists in a practical manner that they can easily analysis the difficult data within very short period and more accurate way that was previously considered impossible. Here we will study the intersection between deep learning with computer image analysis that are pertinent to life scientists. We will study the principles of CNN, image classification, different CNN models and lastly the Flask application. Later we will implement deep learning with annotation of training data, selection of data and also present the training a range of neural network architectures and finally the deploying solutions. Here will use four different optimized neural network models named ResNet50, Inception_v3, Xception and VGG-19 to find out among these which will give the most accurate and precise results. The major purpose of our research is to find the model with the best accuracy for future use. From our findings we see that the Xception model has the highest accuracy among the other models we used. The conclusion points to the importance of implementing a proper model for detection of diseases using images and DL for progress in health technology and future researches.