Audio Description Project

A CLI based audio description generation tool built on Python leveraging the power of Ollama, Whisper, CLIP, Coqui TTS and FFMPEG.

Installation

Installation is written for MacOS

1. Install python 3.14
2. Install FFMPEG and Ollama and espeak
   • Run brew install ffmpeg for video editing
   • Run brew install espeak for some AI Text-To-Speech
   • Install Ollama from their site https://ollama.com/download
3. Run ollama pull gemma3:12b and ollama pull nomic-embed-text
4. Run python3 -m venv ./venv && source ./venv/bin/activate
   • Every time you start a new terminal for the project run source ./venv/bin/activate
   • Python Environments will run that for you
5. Run pip3 install -r requirements.txt
6. Run python3 ./describe_video.py --input ./my_video.mp4 --output ./my_video_script.txt to generate a video script file
7. Then run python3 ./process_video.py --input_video ./my_video.mp4 --input_text ./my_video_script.txt --output ./my_video_audio_description.mp4

Pylint

To run pylint, runpylint ./describe_video.py. Pylint configuration is located in ./pyproject.toml