SilentSync-AI

SilentSync AI project bridges communication gap for peoples having hearing and speech impairments by creating a sign language recognition system. This system converts hand gestures into Marathi audio which helps enhancing communication between hearing and speech impaired persons. The project involves three main stages namely data collection, training and testing.The dataset is created mainly using OpenCv which captures hand gestures in ISL Then mediapipe extracts 21 key points for each hand which forms a labeled dataset for training. The system uses tools like CNN, Tensorflow, and Google APIs for interpretation and translation. Finally results show an 95.49\% accuracy in interpreting and recognizing hand gestures. The web interface, developed with HTML,CSS and Django ensures real-time accessibility. Voice synthesis using pygame and gTTs enhances the communication process.Languages support included in future work which allows users to define custom gestures, and developing a mobile app for easier use. SilentSync AI created an inclusive approach to sign language communication, which contributes to a more accessible platform for communication.