communication barriers, Sign Language, Deaf community, LSTM, Communication 

Rwanda's Deaf community faces significant communication barriers that limit their integration into society. Current solutions, like the Rwandan Sign Language (RSL) dictionary, are often costly and inaccessible. To address this, a new machine learning-based RSL translation solution has been developed. This system utilizes Long Short-Term Memory (LSTM) networks and Mediapipe to process real-time video input, accurately interpreting gestures with a 98% test accuracy. By integrating this technology into a user-friendly web application, the project aims to facilitate real-time sign translation and foster inclusive communication. Initial results are promising, with the model performing well in real-world testing. Future plans include expanding the dataset and integrating the solution into a mobile app to further improve user accessibility, ultimately empowering the Deaf community's full participation in Rwandan society. 

Final Year Capstone