I want to build an Android application that allows a deaf-and-mute user to communicate effortlessly with hearing individuals through real-time sign-language translation. The core idea is simple: the camera captures the user’s hand gestures, the app recognises the sign, and instantly converts it into readable text (and optionally speech) on screen. In the opposite direction, typed or spoken words from a hearing person should appear as clear, large-format text for the deaf user. Scope • End-to-end Android solution: intuitive interface, camera feed, translation engine, and clean text display. • Real-time or near-real-time sign recognition; latency under two seconds is the target. • Model training can rely on existing datasets, but the final APK must ship with an optimised on-device model (TensorFlow Lite, ML Kit, or similar) so the main features work offline. • Include a lightweight conversation log the user can save or clear at will. • Accessibility first: high-contrast colours, scalable fonts, vibration cues for status updates. Deliverables 1. Signed APK ready for installation and a full, clean, well-commented source code repository. 2. Documentation that explains build steps, model retraining procedure, and how to extend the sign vocabulary. 3. A short demo video showcasing start-up, signing, recognition, and text output in real time. 4. Acceptance test sheet showing at least 90 % accuracy on a predefined sample of common signs. I will handle Google Play listing later; for now I simply need the working app and instructions so I can test on multiple Android devices (API 24+). If you have prior experience in computer-vision, gesture recognition, or accessibility apps, that will make integration smoother, but well-structured, easy-to-maintain code matters most.