Smart Translation For Physically Challenged People

This paper introduces an AI-driven translation platform developed to enhance communication for individuals with speech and hearing disabilities.The solution integrates advanced technologies such as Natural Language Processing (NLP), gesture recognition using MediaPipe, and Artificial Neural Networks (ANNs) to enable seamless two-way interaction.It supports real-time conversion of spoken language into Indian Sign Language (ISL) through animated GIFs, while also interpreting gestures captured via webcam into coherent text.The system is deployed using a Django-based web application, ensuring both usability and scalability.Unlike conventional tools, this integrated solution enables dynamic, real-time, bidirectional communication, making it highly applicable in fields like education, healthcare, and public services.Experimental evaluations indicate the system achieves an 85% accuracy rate in recognizing gestures and a 92% accuracy rate for speech-to-sign translation, marking a notable advancement over existing approaches