Empowering Speech-Impaired Individuals: A Hand Gesture to Text and Voice Conversion System

This paper introduces an innovative system for converting hand gestures into text and voice, aimed at assisting individuals with speech disabilities. Utilizing the power of Convolutional Neural Networks (CNN), our system achieves an exceptional recognition accuracy of over 99% for hand gestures corresponding to the 26 letters of the alphabet. The model is trained using a comprehensive dataset from Kaggle, ensuring accurate translation of gestures into readable text. To accommodate various user needs, we've incorporated a Graphical User Interface (GUI) built with key libraries such as OpenCV and Keras. This easy-to-use interface enables smooth user interaction and offers real-time word suggestions, making it easier to form sentences. Furthermore, to support users with visual impairments, we have added an option to listen to the generated sentence, promoting inclusivity. With its high precision, user-friendly design, and accessibility features, our Hand Gesture to Text and Voice Conversion System not only showcases the capabilities of CNN-based models but also emphasizes our dedication to creating accessible and inclusive technological solutions.