Hand Kinesics in Indian Sign Language using NLP Techniques with SVM Based Polarity

Abstract

With the advent of new technology every year, human beings continue to make clever innovations to benefit not only themselves but also those with some kind of impairment. Communication is carried out by talking to each other for regular people, but people who are deaf interact with each other through sign language. Taking this problem into account, we are proposing a methodology that allows to ease the communication with each other by translating speech into sign language. This paper explains a methodology that translates speech into the corresponding Indian Sign Language (ISL). In India, it is spoken in almost 28 different languages. So, language has always been a problem. Thus, we have come with a project just for India in which the person can communicate with the app in any Indian language they know, and it will convert it into Indian Sign Language. This is applicable to not just literate but also illiterate people across India. The idea is to take the speech input and translate to text, which will then undergo text-pre-processing using NLP for better analysis and will be connected to the HamNoSys data for the generation of sign languages. The polarity detection will also be included. It is implemented using the SVM algorithm for Sentimental Analysis. Thus, the main objective of this project is to develop a useful project which can be used to capture the whole vocabulary of Indian Sign Language (ISL) and provide access to information and services to mute people in ISL.