Takes a .pptx file as input, then uses GPT-4 vision + any embedded speaker notes to write a transcript for the whole presentation. Finally it uses OpenAI's TTS to create a believable audio track, and compile the whole slide show as a video.

• To get started, clone this repo

git clone https://github.com/chaonan99/ppt_presenter.git
cd ppt_presenter

• Install required packages

pip install -r requirements.txt

• Install other dependencies

• ffmpeg
• libreoffice
• poppler

• Watch the video example/test.mp4