Deep Learning Model for Image Classification Using Convolutional Neural Network

In many signal and image applications, deep learning has emerged as a crucial field of machine learning. It is a complicated process that depends on various elements. The application of one of the powerful deep learning algorithms, the Convolutional Neural Network (CNN), is picture categorization. The main goal of the work presented in this study is to detect translation, scaling, and other types of distorted and invariant image distortion with the help of deep convolution neural networks. This CNN model will be trained to detect different Hotel Room images and classify them based on trained data. A Flask application will be built to detect the image uploaded by the customer. In the background of this Flask application, the hotel room image classification model (CNN) will be running to detect the images. This Flask Server will be deployed on AWS (Amazon Web Services) EC2 (Elastic Compute Cloud) Machine so that users from all over the world can access it using either the Internet Protocol (IP) address of the machine or by using the DNS (Domain Name System) name assigned to that corresponding IP address. The researcher used the VGG16 (Visual Geometry Group) model of CNN architecture to build the model. For visualizing the accuracy of the built model, the Matplotlib library built on NumPy arrays is used. The experimental result analysis based on the graphical representation and quality metrics shows that the CNN algorithm provides moderately better classification accuracy for all tested datasets.