Published Paper Details:

In the modern digital era, effective communication remains a significant challenge for individuals with hearing and speech impairments. Traditional sign language communication often requires trained interpreters, which limits accessibility and real-time interaction with the general population. Although several assistive technologies have been developed, many existing solutions suffer from limited accuracy, lack real-time processing, dependency on internet connectivity, and poor user adaptability. This paper proposes an AI Based Sign Language Recognition and Language Translator designed to bridge the communication gap between sign language users and non-signers. The proposed system utilizes computer vision and deep learning techniques to recognize hand gestures captured through a webcam and convert them into meaningful text and speech in real time. MediaPipe Hand Tracking is employed to extract hand landmark features, which are classified using a custom-trained neural network model. To improve reliability, a confidence-based filtering mechanism is implemented to reduce misclassification of gestures. Additionally, the system integrates an offline speech-to-text module and a multilingual translation component, enabling seamless bidirectional communication across different languages. A Streamlit-based interactive user interface provides real-time visualization, gesture training support, and smooth user interaction. The proposed solution offers a lightweight, scalable, and practical approach for assistive communication, accessibility technologies, and inclusive educational environments.