YOLOv5-Based System for American Sign Language Interpretation

Abstract

Sign language is a vital form of communication for individuals with hearing and speech impairments. This project presents a real-time sign language recognition system using the YOLOv5 deep learning model. The proposed system efficiently identifies hand gestures from video input, converting them into readable text and speech output. It uses a CNN-based architecture that processes grayscale images for reduced computational cost without compromising accuracy. The system was trained and validated on robust datasets such as the Massey University Dataset and the ASL Alphabet Dataset, achieving over 99% accuracy. This solution enables inclusive communication, supporting human-computer interaction, assistive technologies, and real-world applications in education, healthcare, and social settings by providing a fast, scalable, and user-friendly gesture recognition platform.