michaelchinaka/shona-realtime-STT

# 🎙️ Shona Live Transcription A real-time speech-to-text application with intelligent English translation, specifically designed for the Shona language. This application provides live transcription with a beautiful, modern React frontend and a powerful FastAPI backend. ## ✨ Features - **Real-time Speech Recognition**: Live transcription of Shona speech using Google Cloud Speech-to-Text - **Intelligent Translation**: Automatic translation from Shona to English using Google Translate - **Modern UI**: Beautiful React frontend with real-time microphone level indicators - **Presentation Mode**: Full-screen mode perfect for presentations and demos - **WebSocket Communication**: Low-latency real-time communication between frontend and backend - **Audio Processing**: Modern AudioWorkletNode-based audio processing for reliable microphone input - **Confidence Scoring**: Display confidence levels for transcriptions - **Conversation History**: Keep track of recent transcriptions ## 🏗️ Architecture ### Backend (FastAPI) - **FastAPI**: Modern, fast web framework for building APIs - **WebSocket**: Real-time bidirectional communication - **Google Cloud Speech-to-Text**: Shona language ASR (Automatic Speech Recognition) - **Google Translate**: Translation services - **Audio Processing**: Real-time PCM audio processing ### Frontend (React + Next.js) - **React 18**: Modern React with hooks and functional components - **Next.js 14**: Full-stack React framework - **TypeScript**: Type-safe development - **Tailwind CSS**: Utility-first CSS framework - **Framer Motion**: Smooth animations and transitions - **AudioWorkletNode**: Modern Web Audio API for reliable audio processing ## 🚀 Quick Start ### Prerequisites - Python 3.8+ - Node.js 18+ - Google Cloud account with Speech-to-Text and Translate APIs enabled - Google Cloud credentials JSON file ### Backend Setup 1. **Clone and navigate to the project**: ```bash git clone cd google_shona ``` 2. **Create and activate virtual envir …