Voice Bridge: Sign Language Translation to Text

The primary mode of communication is speech; however, this posses significant challenges in individuals having impaired speaking or/and hearing abilities. This work introduces a technique for recognizing Indian sign language. (ISLRS) designed to capture signs in real-time, accurately detect them, and convert them into text. Leveraging sophisticated technologies like natural language, deep learning, and computer vision processing, the ISLRS aims to mitigate communication problems encountered by deaf people that promote inclusivity in diverse settings. A deep learning (DL) model which can identify and detect words from an individual's gestures is presented in our work. Long-term short-term memory (LSTM) networks were utilized to identify signs from video frames of isolated Indian Sign Language (ISL). The suggested model, which has three LSTM layers, attains a noteworthy accuracy of around 92.5demonstrates its effectiveness in facilitating accurate and reliable sign language recognition.