Sign Language Translation with fusion of Emotion Detection
Ashwin Acharya, Navin Patil, Utkarsh Pathak, Sumedha Bhagwat · 2024
Sign languages play a crucial role in fostering natural interaction, breaking down communication barriers between the hearing impaired and society. However, recognizing words in sign language poses significant challenges, particularly when gestures for multiple words are similar. Additionally, the rapid transition between gestures during communication complicates the creation of coherent sentences based on recognized words. To address these challenges, we propose the Real-Time Sign Language Recognition and Emotion Detection (RTSLRED) model, leveraging MediaPipe’s holistic pipeline and LSTM network. Specifically focusing on American Sign Language (ASL), our model aims to enhance accessibility and inclusivity for the hearing-impaired community. Our results indicate an impressive accuracy of 91.05%, showcasing the model’s effectiveness in conveying emotions and sentences in sign language. This innovation holds promise for empowering the hearing impaired and promoting inclusive communication practices.