Term 252

Smart Glove for Translating Sign Language into Speech

Project Type: Self-Initiated

Project Description

This project presents the design and development of a wearable smart glove that enables real-time two-way communication between sign language users and non-signers. The system captures hand and finger movements using embedded flex and motion sensors, processes the signals through a low-power microcontroller, and applies a machine learning–based classification algorithm to recognize at least 10 commonly used sign language gestures. The system achieves a response time of less than 5 seconds, operates continuously for approximately 6 hours, and maintains reliable performance within a communication range of at least 5 meters. Recognized gestures are converted into spoken output through an integrated wrist-mounted speaker, while speech input is translated into text for the user through a connected interface. In addition, a dedicated mobile application provides a camera-based translation mode as a backup solution in case of hardware limitations or failures, ensuring continuous and reliable communication. The design emphasizes real-time performance, energy-efficient operation, user comfort, and mechanical reliability to ensure practical daily use. The outcome is a functional AI-based assistive prototype that demonstrates reliable gesture-to-speech and speech-to-text translation, providing a scalable foundation for future improvements in accuracy, expanded vocabulary, and enhanced user interaction. Ultimately, the system aims to reduce communication barriers and promote inclusive, independent communication in everyday environments.