Audio Transcription with OpenAI Whisper for Education and Brain Sciences Research Lab

A Python project that uses OpenAI's Whisper model to transcribe children's reading testing audio.

Setup

1. Clone this repository
2. Install the required dependencies:

pip install -r requirements.txt

3. Create a .env file with your OpenAI API key:

OPENAI_API_KEY=your_api_key_here

4. Place your audio files in the audio_files/ directory

Usage

Transcription

You can place your audio files in the audio_files/ directory and then simply refer to them by filename:

python transcribe.py filename.mp3

Or provide a full path to an audio file anywhere on your system:

python transcribe.py /path/to/your/audio/file.mp3

You can also specify an output file for the transcription:

python transcribe.py filename.mp3 -o transcription.txt

To use a different audio files directory:

python transcribe.py filename.mp3 -d your_audio_directory

Audio Format Conversion

This project includes a tool to convert various audio formats (WMA, DSS) to MP3:

Convert a single file:

python wma.py your_file.wma
python wma.py your_file.dss

Convert a single file with custom output path:

python wma.py your_file.wma -o output.mp3

Convert all supported files in a directory:

python wma.py your_directory/

Convert all supported files in a directory with a custom output directory:

python wma.py your_directory/ -o output_directory/

Convert all files including in subdirectories:

python wma.py your_directory/ -r

Specify a custom bitrate for the MP3 output:

python wma.py your_file.wma -b 320k

Convert only specific file types:

python wma.py your_directory/ -t wma,dss,wav

Interactive Usage

Run the interactive sample script:

python sample_usage.py

Features

• Transcribes audio files using OpenAI's Whisper model
• Automatically converts audio to the required MP3 format using pydub
• Supports saving transcriptions to a text file
• Environment variables for secure API key storage
• Dedicated audio_files/ directory for organizing your audio files
• Audio format conversion utility (WMA, DSS to MP3)

Supported File Formats

The transcription tool can handle various audio formats including:

• MP3
• WAV
• FLAC
• OGG
• WMA (will be automatically converted to MP3)
• DSS (will be automatically converted to MP3)
• and more (supported by pydub/ffmpeg)