Published Paper Details:

This project presents the development of an AI-powered Pronunciation Enhancement Application designed to help users, particularly Indian English speakers, improve their spoken English through real-time feedback and personalized insights. This project uses an Indian-English fine-tuned Wav2Vec2 for live speech recognition , performs phoneme-level correctness , word level correctness also the system identifies pronunciation errors and highlights areas for improvement by providing insights . The application is built using a Flask backend and a React.js frontend, ensuring a responsive and interactive user experience. Key features include real-time transcription, audio comparison with ideal pronunciations using ElevenLabs TTS, multilingual support, and dynamically adjusted practice difficulty. The backend also delivers detailed performance metrics and session insights, which are visually interpreted on the client side. By integrating advanced AI models with intuitive web technologies, this system offers an engaging and accessible tool for users seeking to refine their pronunciation skills. The project was successfully tested meeting its core aspects , which is to be a enhanced web application for pronunciation improvement for non-native speakers.