Speech-AI-Transcribe Tool

About

Speech-AI-Transcribe is a powerful tool for converting spoken language to written text directly in the browser. It uses advanced speech recognition technology and is easy to integrate into many types of projects. This tool is especially useful for those needing transcription of media files or real-time audio streams.

Examples

Video On Demand (VOD) Content or Streams

In the main example, you can see how to create movies where the word being said is highlighted.

Additional Examples

How to Install

Preparation Steps

First, make sure Node.js is installed on your computer.

Next, add the module to your project by running:

npm install speech-ai-transcribe

How to Use It

The Speech-AI-Transcribe offers two main ways to work: FileTranscriber for files and StreamTranscriber for live audio.

Server Setup

Make sure your web server is set to handle files correctly:

• "Cross-Origin-Embedder-Policy": "require-corp"
• "Cross-Origin-Opener-Policy": "same-origin"

NPM Way

If you use npm, install the package and move needed files to your project folder:

npm install --save speech-ai-transcribe
cp node_modules/speech-ai-transcribe/dist/* /your/project

Manual Way

If not using a package manager, download and copy the dist/ files from the repository to your project manually:

<script type="importmap">
{
  "imports": {
    "speech-ai-transcribe": "/path/to/speech-ai-transcribe.js"
  }
}
</script>
<script type="module">
  import { FileTranscriber } from "speech-ai-transcribe";
  ...
</script>

Example to Follow

To make your audio or video files into text, setup the transcriber with model and worker path:

import { FileTranscriber } from "speech-ai-transcribe";

const myTranscriber = new FileTranscriber({
  model: "/path/to/model.bin",
  workerPath: "/path/to/worker"
});

await myTranscriber.init();

const result = await myTranscriber.transcribe("/path/to/file.mp3");
console.log(result);

You get a JSON object with text parts, times, and chance scores.

How to Develop

For those interested in contributing or tweaking:

git clone https://github.com/aresobus/speech-ai-transcribe
cd speech-ai-transcribe
npm install
npm run dev

Open http://localhost:9876/examples/index.html to see your changes in a local development setup.

Testing

Run unit tests and functional tests with:

npm run test:unit
npm run test:e2e

Create TypeScript definitions from JSDoc:

npm run generate-types

Thanks

Big thanks to many open source projects and contributors that help make Speech-AI-Transcribe possible.

License

This project is shared under the MIT License, allowing everyone to use, change, and share within the license terms.