International Research Journal of Education and Technology

Sign language is an essential means of communication for individuals who are deaf or hard of hearing, enabling them to interact effectively with others. However, the lack of widespread understanding of sign language among the general population creates significant communication barriers. This results in difficulties in accessing essential services such as healthcare, education, and employment. While traditional sign language recognition systems have made progress in bridging this gap, many rely on specialized hardware, such as gloves or motion sensors, limiting their practicality and scalability in real-world settings. This research presents a deep learning-based approach to sign language recognition and speech-to-sign language conversion using computer vision techniques. The system integrates MediaPipe for real-time static gesture recognition and Long Short-Term Memory (LSTM) networks for dynamic sequence modeling, allowing for more accurate sign language interpretation. By leveraging publicly available datasets, including Indian Sign Language (ISL) and American Sign Language (ASL), the proposed system is designed to be scalable, adaptable, and capable of supporting multiple sign languages. By addressing these challenges, this research aims to create an inclusive, technology-driven solution that fosters better social integration and accessibility for individuals with hearing impairments. The proposed system has potential applications in various real-world scenarios, including educational institutions, workplaces, and public services, contributing to a more inclusive society where sign language users can communicate effortlessly with the broader population.

Key Words:   Sign language recognition, deep learning, MediaPipe, LSTM, real-time communication, accessibility, Indian Sign Language (ISL), computer vision, gesture recognition.