ai-speech-text-translation

Demo React/Tailwind app with AI speech-to-text and translation capabilities. This application uses browser-based machine learning models to convert English speech to text and translate it to Dutch.

Features

• Speech-to-text conversion using Whisper AI model
• English to Dutch translation using Opus MT model
• Real-time audio recording and transcription
• Modern UI built with Tailwind CSS
• Fully client-side processing - no server required

Technologies

• React 19
• Tailwind CSS
• Hugging Face Transformers.js
• Web Audio API

Installation

First, install all dependencies:

npm install

Development

To run the development server:

npm start

This will start the app in development mode. Open http://localhost:3000 to view it in your browser.

Production Build

To create a production build:

npm run build

This will create an optimized build in the build folder.

Important Notes

• The browser will request microphone permissions for audio recording.
• ML models are downloaded when first visiting the page, which may take some time.
• All processing happens locally in the browser - no audio/text is sent to external servers.
• This readme was written by Claude 3.5 Sonnet, based on the contents of this repository.