Dynamic Sign Language Detection: A Web based Approach for Real Time Adaptability

This paper introduces a pioneering framework for sign language detection, leveraging a dynamic web-based interface and the K-Nearest Neighbors (KNN) algorithm. Departing from traditional static repositories, our system enables real-time recognition and seamless adaptation to novel gestures. By integrating user-generated gestures and employing self-training mechanisms, our model achieves high accuracy levels across diverse gesture classes. Through rigorous experimental validation, including dataset curation and performance metrics evaluation, we demonstrate the efficacy and reliability of our approach. This innovative solution holds promise for enhancing communication accessibility among individuals with auditory impairments, fostering inclusivity and understanding in society.