Hands in Harmony: Empowering Communication Through Translation

Over the years, sign language has developed to be a remarkable advancement. Unfortunately, there are specific effects associated with this language. When speaking with speech disabilities or hard-of-hearing people, not everyone knows how to decode this sign language. This study investigates the notion of voice-to-Telugu text, sign language conversion systems, and sign language-to-Tamil text conversion systems to facilitate seamless communication between hearing and hard-of-hearing people. Using automated speech recognition (ASR) algorithms, it transcribes spoken words into text and produces sign language animations or gifs using computer vision methods. Similarly, sign language conversion systems use BiLSTM neural networks and media pipe holistic to understand sign language motions in real-time and translate them into Tamil. The proposed work stores the weights and uses the confusion matrix accuracy to assess the model. By overcoming different communication challenges, this well-designed architecture offers a practical and comprehensive solution for deaf and hard-of-hearing society.