Real-Time Sign Language Translation System Using QCNN for Enhanced Accessibility

Sign language functions as the main form of communication which deaf and hard-to-hear people require to interact effectively. The inadequate understanding of sign language prevents effective communication from happening within society. This work proposes a real-time Sign language detection system using Quantum Convolutional Neural networks (QCNNs) deployed in a Raspberry pi to enhance recognition accuracy and operational speed. The implemented model demonstrated training and testing performance at 98.34% accuracy whereas real-time operation reached 97.84% accuracy. The performance of feature extraction and classification improves when QCNNs merge classical and quantum computing elements into their operation. Live hand gestures detected by the model through camera input produce textual outputs that are converted to spoken commands which use TTS (Text-to-Speech). The system combines various communication tools to create accessible sign language communication that improves daily interaction quality.