Speech to Sign Language Translation for Indian Languages
Jashwanth Peguda, V Sai Sriharsha Santosh, Y Vijayalata, R. Deepa, Vaddi Mounish
Abstract
Hearing-impaired people and mute people face a lot of difficulty in communication while interacting with others in society. It may reduce their self-confidence and might make them feel isolated from others. Sign language acts as a communication medium between deaf people and ordinary people. Many technologies are used to convert the text to American Sign Language. There is a limited amount of research done on Indian Sign Language and is widely used by deaf people in India. This research aims at conversion of speech to Indian sign language for six Indian regional languages Telugu, Hindi, Malayalam, Marathi, Kannada and Tamil. The proposed model takes speech as the input and displays a sequence of corresponding gestures as the output. It involves speech recognition using Wavelet-based MFCC with GMM, text translation using LSTM and mapping the text with the sign language.