Sign Language Translator using Accelerometers and Deep Learning

This project introduces a real-time ASL hand gesture recognition system using a smart glove equipped with five MPU6050 accelerometers. Data from the sensors is transmitted via a TCA9548A multiplexer to a Jetson Nano using I²C. A C-based program collects motion data, which is converted into scalograms using Continuous Wavelet Transform and classified by a CNN. The system translates gestures into text or speech, offering a reliable, low-latency alternative to camera-based solutions, unaffected by lighting conditions and suitable for individuals who are deaf or hard of hearing.

Project Poster