Assistance to Uzbek people with speaking disabilities using Machine Learning and Sign Language Detection

This paper presents an ongoing project addressing the communication needs of individuals with speaking disabilities, focusing on Uzbek sign language users. Leveraging machine learning, the application detects hand gestures in real-time, translating them into spoken Uzbek using Python, OpenCV, MediaPipe, and scikit-learn. Despite challenges in dataset creation, the Random Forest classifier achieves high accuracy. The current version demonstrates real-time hand gesture capture, accurate sign language recognition, and spoken translation. This work serves as a progress snapshot, with future plans to extend the application for real-time gesture recognition. The insights gained contribute to advancing assistive technologies, offering a promising solution for effective communication through sign language.