IMAGE INSCRIPTION AND INTONATION -A NEURAL NETWORK APPROACH

Abstract

Digital communication technologies have greatly influenced and expanded the way humans interact. The progress of information technology has opened wider opportunities for communication. Social networks have become the modern-day social communities connecting people from different parts of the globe, sharing images and videos on these platforms. By creating virtual communities, digital communication has expanded the scope of communication eliminating barriers. We aim to make further progress in this arena by describing an image in the form of audio to visually impaired people. A certain section of differently abled people is unfortunately isolated from this world. In-order to combat this issue we have come up with a system that describes an image shown in the form of plain text using an encoder-decoder architecture and is integrated with an end-to-end lexical articulator which produces a vocal description of the given image.