Machine Learning Technology to Recognize American Sign Language Alphabet

Recent technological advancements have facilitated the development of a variety of tools and software designed to enhance the quality of life for individuals with hearing impairments. In this research paper, a comprehensive investigation was conducted utilizing three distinct machine learning models to interpret hand gestures representing the American Sign Language (ASL) alphabet. The study employed the Random Forest, Support Vector Machine (SVM), and K-Nearest Neighbors (KNN) models, all trained and tested using a dataset comprising 87,000 images of ASL alphabet hand gestures converted into grayscale pixels. Multiple experiments were performed, involving adjustments to the models' architectural parameters to achieve the highest possible recognition accuracy. The experimental results were exceptional; the Random Forest model achieved an outstanding accuracy rate of 99.55%, the highest among all models. The SVM model achieved an accuracy rate of 99.29%, while the KNN model reached an accuracy rate of 98.69 %.