IMAGE CAPTIONING USING IMAGENET

1S.INIYAN, PRASHANTH MAHESWARI, RAHUL AJITH

124 Views
37 Downloads
Abstract:

There are many cases wherein an image has to be described to people, or a caption is needed for multiple reasons. Giving pre-defined captions for each specific image can be a long and dreary job for a human being when there is an excessive number of images involved. This is where the image captioning system comes into play. In this paper, we explore the mapping between images and their descriptions in a sentence form. It can be useful in creating something that generates natural language which can describe the image in a manner that is understandable by human beings. Making for more human like responses can greatly benefit the human race as many things can be computerized in the near future which takes the tedious work of captioning given images in a large scale off our hands. What is the use for computer generated image captioning? People may need to find out what the object in front of them is, in case it is something that they aren’t acquainted with, or they may want a description of what’s happening in the given image. If the system has a reference that can be used to detect the image, it can be beneficial to the end user. On a large scale, this can be used as a tool that can work as an assistant, potentially connected to a camera or a storage device which contains images for it to work on.

Keywords:

Captioning, Imagenet

Paper Details
Month4
Year2020
Volume24
IssueIssue 6
Pages6653-6664