Automated Gesture Recognition using Deep Learning Model for Visually Challenged People

Individuals with visual impairments face challenges in engaging in tasks involving surroundings, social interactions, and technologies. Moreover, individuals are experiencing challenges in being self-reliant and secure in their everyday activities. The blind people may precisely sense and react to emotions with recognition. The current application needs the integration of face and facial expression detection. Technologies seem far more sophisticated than they were in the past. It is possible to identify the communication of a deaf and visually challenged individual by recording their speech and comparing it to existing datasets, therefore determining their intentions. This research presents a system for recognizing hand gestures and faces using animated pictures and techniques. The hand gesture method identifies skin color and hand convex deformities, while the face recognition system utilizes Haar Cascade Classifiers and LBPH recognizer for identification and authentication. OpenCV is used for execution. The study achieved an accuracy rate of $96.3 \%$ in identifying hand gestures and facial features. The system is automated and operates on an artificial intelligence server.