ENHANCING WEB ACCESSIBILITY USING DEEP CONVOLUTIONAL NETWORKS AND NATURAL LANGUAGE PROCESSING TECHNIQUES

Abstract

Deep neural networks (DNN) and Convolutional neural networks (CNN) are artificial neural networks used for image classification, natural language processing, object detection, and image segmentation. These techniques aid in the friendly usage of websites for people who have some kind of disability making it difficult for them to access. In this study, DNN and CNN were opted and employed to generate captions for the given images using different datasets, and metrics such as, BLEU and WER were used for system evaluation. The study's results revealed promising outcomes, highlighting the efficacy of deep learning techniques in enhancing web accessibility for individuals with visual impairments. The developed system effectively enhances the browsing experience and improves information accessibility for individuals with print impairments by providing precise and descriptive captions for images. These advancements align with the broader objective of enabling intelligent machines through the utilization of natural language processing (NLP) and facilitating linguistic based communication between humans and compute