audio-transcriber-web

日本語のREADMEはこちらから

Using SpeechRecognition of browser, audio from the microphone automatically transcribe and written to file.

• GitHub Pages
• For transcriptions from video and audio files: book000/audio-transcriber-docker

Warning / Disclaimer

The author is not responsible for any problems caused by the user using this project.

Or rather, if you want to use it seriously, use Google Cloud Platform Speech-to-Text.
We do not provide any support.

Feature

• Input the API URL, can send the transcribed text to the specified address automatically.
• With the AutoStart feature turned on, recording will start the moment you visit the page or reload. (beta)
• The API URL and AutoStart settings are stored and maintained in localStorage.
• Shows alternatives to recognize sentences and diffs by hovering the results.

Requirements

• Chrome 33+ / Edge 79+ (Can I use SpeechRecognition?)
• Python 3.6+
• requirements.txt: fastapi, uvicorn

Installation & Usage

1. Git Clone

git clone https://github.com/book000/audio-transcriber-web.git

2. Install packages from requirements.txt

pip3 install -U -r requirements.txt

3. Run main.py --open-browser

python3 main.py --open-browser

4. Type http://localhost:8080/api in the API URL form

5. Click the Start recording

If you turn on AutoStart, recording will start the moment you access the page.

Notes

• It also works if you put api.php on a server somewhere and set the address to it to an API URL.

License

The license for this project is MIT License.

This project contains several libraries. The library licenses are as follows.

• axios MIT License, Copyright (c) 2014-present Matt Zabriskie
• jsdifflib BSD license, 2007 - 2011 Chas Emerick