A Comprehensive Neural Network Model for Colorization and Captioning of Images Along with IoT Deployment

The convergence of artificial neural networks and the internet of things (IoT) has gained popularity in the field of computer science research. In this work, an efficient neural network model for the image colorization problem is proposed along with deploying these models to the remote system using IoT deployment tools. Further, this work proposed two convolution neural network models namely the Alpha model and Beta model towards solving the image colorization of the grayscale format. An efficient combination of models is proposed and analyzed such that the loss rate is minimized as?~?0.005. Next, an efficient model for solving image captioning is proposed based on the bi-directional long short term memory model. Finally, the work discusses the merits and demerits of deploying the neural network model using the AWS Greengrass and Docker IoT environment on remote systems.