Accessify: An ML Powered Application to Provide Accessible Images on Web Sites

Abstract

Web Accessibility remains a major concern of the rapidly growing Internet community. Accessing online content is crucial for many people as they engage in a variety of tasks every day. Content can be in any form, such as audio, video, images and text. A major portion of a website's viewport generally contains images, and there are set standards & guidelines for providing information about these images while creating image elements on a webpage. Yet, most websites do not strictly follow accessibility guidelines and, as a result, either a website contains no descriptions of images or the descriptions provided are not sufficient to convey the meaning to visually impaired people who use screen readers for accessing web content. In this paper, we propose a solution that provides alternate image descriptions to all images on a website with help of Machine Learning in a completely unobtrusive environment. The application does not require any setup and can be used on static as well as dynamic websites.