HAND GESTURE TO VOICE CONVERSION USING ARTIFICIAL INTELLIGENCE

Abstract

Communication forms one of the basics of interaction among people but becomes a significant problem for deaf and mute individuals. Conventional approaches, such as sign language and lip reading, tend to be restricted regarding availability and precision. The proposed project will close this communication gap with the help of an image-processing-based hand gesture recognition system. This system allows hand signals to be captured by a webcam, converted to text, and then to speech, providing a readily understandable input/output medium. This work captures the hand gestures, preprocesses the model to be invariant to lighting and background variations, and tests with measures such as IOU (Intersection Over Union), MAP (Mean Average Precision), MAE (Mean Absolute Error), and RMSE (Root Mean Square Error). Further, the proposed approach provides better results for hand gestures using image processing by employing deep-learning algorithms for feature extraction and real-time recognition.