Mozilla DeepSpeech with Vaadin

How it works:

This component is based on Mozilla DeepSpeech. More specifically on their web_microphone_websocket example. This implementation asks for microphone access and if granted, it records voice, sends it to a node.js server, which again sends the recognized text back. This text can then be processed in your Java code.

How to use:

1. Get the Node.js server up and running. Here are the installation instructions.
2. Grab yourself a copy of the downsampling_worker.js file. Place this file in the projectRoot/src/main/webapp/ folder. (Or wherever the file will become available under yourdomain.com/downsampling_worker.js)
3. Go to your projects root directory and run npm add socket.io
4. Add the maven dependency to your pom.xml
5. Use the component in your Vaadin application (see the code samples).

More information:

For more Information about this project see the GitHub repository.