Image Caption Generator

Abstract

Image captioning, also defined as describing the image, has consistently sparked the curiosity of expert system researchers and accurate description of an image has been a significant task.Image caption generator involves describing the characteristics, attributes of the image.It has a plenty of applications in the field of Robotic vision, story-telling from album uploads, business and many more.For instance, it can be used in Image segmentation as used by Google Photos and its application can also be extended to video frames.It has grown to become one of the most prevalent tools in the contemporary period.This paper aims in employing computer vision and machine translation for captioning the image.It involves recognizing the objects, actions, attributes in an image and identify the relation between the objects and the generated descriptions.Most of them use encoder-decoder framework, where the image, which is given as input, is encoded to an intermediary representation of the image's information and then decoded into a series of descriptions and descriptive text.The dataset employed for the same is Flickr8k dataset and the programming language is python.The project involves developing an app that takes an input image, extract features, and generate accurate descriptions, using Flutter.It has an immense potential in helping the visually impaired.It helps in automating the job of radiologists.