A Conversion from Sign Language to Text & Speech using Machine Learning

Language is the medium through which people communicate, which is a common need for humans. Most of the people are able to speak and listen, and they communicate using various languages. The use of signals by people with hearing impairments is common. Those who have hearing loss in America prefer to communicate using American Sign Language (ASL). It becomes essential to use PC-based ASL translators to make conversation between hearing-impaired and hearing people easier and less complicated. This project aims to develop a vision-based application that provides sign language to text translation, facilitating conversation between signers and non-signers. Video sequences are taken in consideration for the suggested model to extract material and contiguous features. Then, to identify spatial characteristics, we employ Inception, a CNN (Convolutional Neural Network). Then, to learn on temporal features, we employ an RNN (Recurrent Neural Network). The American sign language dataset was utilized.