Hand Gesture Recognition for Indian Sign Language Communication

Abstract: Effective communication between people with hearing impairments and the wider population continues to pose challenges in education, work, and daily life. Indian Sign Language (ISL) serves as a crucial communication tool for individuals with hearing and speech difficulties in India. This study proposes an automated system for recognizing ISL gestures, utilizing computer vision and machine learning techniques to translate hand movements into text or speech in real time. The system captures gestures through a camera and applies preprocessing steps such as background subtraction, hand detection, and feature extraction to accurately interpret hand movements. A Convolutional Neural Network (CNN) is trained on a dataset of ISL signs to classify gestures with high precision. In addition, the system can process dynamic gestures, which involve a sequence of movements, enabling comprehensive recognition of complex signs. By converting recognized gestures into text and speech instantly, the system provides a real-time interface that enhances communication for hearing-impaired individuals in educational, social, and professional settings. Experiments demonstrate that the system performs reliably under varying lighting conditions, hand orientations, and backgrounds. Overall, this work contributes to greater accessibility and inclusivity for the hearing-impaired community in India.