SimpleTranscriberWebApp

A simple web application to transcribe audio files to text using OpenAI's Whisper model.

Motivation

Good audio-to-text converters are expensive. Open AI released Whisper AI which is a state-of-the-art audio-to-text model. However, using Whisper requires some implementation. This repository aims to provide a coding-free state-of-the-art transcription tool free of charge.

Features

• Optimize for speed or precision by choosing the model size (tiny, base, small, medium, large)
• Support for multiple audio or video formats (flac, m4a, mp3, mp4, mpeg, mpga, oga, ogg, wav, webm)
• Option to download transcribed text in various formats: (.txt, .docx, .pdf, .html, .json)
• Automatic language detection for transcription. Other Languages can be prespecified for faster transcription
• Built-in translation. Just select your target language and the tool will translate the audio as well

Installation

Clone the repository and install the requirements:

git clone https://github.com/henningheyen/SimpleTranscriberWebApp.git
cd SimpleTranscriberWebApp
pip install -r requirements.txt

Getting Started

Once you have all require packages installed the app can be lauched with

python app.py

Then vist http://127.0.0.1:5000 in your Browser.

Make sure your audio files are placed in the uploads folder. Your transcripted files will be located in the transcripts folder or can be downloaded from the WebApp directly. You can find some example audio files and its transcripts in this repository.