SenseGlove 

SenseGlove is a real-time Sinhala sign language interpreter designed to be user-friendly, portable, and affordable, aiming to bridge the communication gap between hearing-impaired and hearing communities in Sri Lanka. It uses custom 3D-printed components, with finger movements tracked by potentiometers and hand movements in 3D space monitored by an MPU 6050 sensor. An ESP32 microcontroller transmits sensor data to a PC via Bluetooth serial communication.

A gesture detection model, based on Long Short-Term Memory (LSTM) techniques, was trained to recognize six gestures with 100% accuracy and an inference time of just 120ms. The system also integrates the OpenAI API to generate meaningful sentences based on detected gestures.

The technology is intended to evolve further to support and empower the hearing-impaired and deaf communities in Sri Lanka and globally.