Sign Language Recognition System for Service-Oriented Environment

This paper presents a real-time sign language recognition and translation system, designed to improve communication for the deaf and hard-of-hearing community. The system uses YOLOv5 and Convolutional Neural Networks (CNNs) to translate American Sign Language (ASL) gestures into text. Trained on 2369 images and validated with 342 images covering 49 ASL signs, the model achieved high accuracy in real-time tests. Intended for service-oriented environments, this system enhances accessibility and inclusivity by facilitating communication between sign language users and those unfamiliar with ASL.