Colorizing and Captioning Images Using Deep Learning Models and Deploying Them Via IoT Deployment Tools

Abstract

Neural networks and IoT are some top fields of research in computer science nowadays. Inspired by this, this article works on using and creating an efficient neural networks model for colorizing images and transports them to remote systems through IoT deployment tools. This article develops two models, Alpha and Beta, for the colorization of the greyscale images. Efficient models are developed to lessen the loss rate to around 0.005. Further, it also develops an efficient model for the captioning of an image. The paper then describes the use of tools like AWS Greengrass and Docker for the deployment of different neural networks models, providing a comparative analysis among them, combining neural networks with IoT deployment tools.