Comparative Study on Image Captioning

Abstract

Image captioning is a technique to generate a correct description of an image with proper structure. Now a days, Machines can describe the images and convert it into suitable language which is grammatically true. In this paper, we reviewed the existing image captioning techniques in details, divided into different categories, datasets, and results. We are focusing on those methods, which are correct grammatically and syntactically and related to deep learning (Neural network). We have also compared and reviewed the different datasets like COCO, flicker8K, etc., and results of different image captioning techniques.