Automatic Image Captioning Methods

2020 VOLUME-8 ISSUE-10, AUGUST 2019, REGULAR ISSUE  
A language known to humans is a natural language. In computer science it is the most challenging task to make the computers understand the natural languages and generating caption automatically from the given image. While a lot of work has been done, the total solution to this problem has been demonstrated daunting so far. Image captioning is a crucial job involving linguistic image understanding and the ability to generate interpretation of sentences with proper and accurate structure. It
more » ... res expertise in Image processing and natural language processing. The publishers suggest in this practice a system using the multilayer Convolutional Neural Network (CNN) to generate language describing the images and Long Short Term Memory (LSTM) to concisely frame relevant phrases using the driven keywords. We aim in this article to provide a brief overview of current methods and algorithms of image captioning using deep learning. We also address datasets and measurement criteria widely used for the same.
doi:10.35940/ijitee.g1013.0597s20 fatcat:ru7qyejarrgkxayzlmqtxw4sci