AI Dubbing System
Built an end-to-end AI pipeline that automatically dubs English videos into 5 languages using Whisper, Helsinki-NLP and gTTS. Implemented audio time-stretching using Librosa to synchronize dubbed speech with original video timestamps. Deployed live web application on Streamlit with video upload, language selection and download functionality. Integrated 4 AI models in a single pipeline covering speech recognition, translation and voice synthesis.