This Python script extracts text and images from a PowerPoint presentation (.pptx file), summarizes the text using a Hugging Face transformer model (facebook/bart-large-cnn), and exports the summarized content to a markdown file compatible with Obsidian note-taking software.
- Text Extraction: Extracts and cleans text from each slide in the
.pptxfile. - Image Extraction: Saves images from the slides into a specified directory.
- Text Summarization: Summarizes the text content of each slide using a pre-trained model from Hugging Face.
- Markdown Export: Converts the slide data into a markdown file, formatting titles, summaries, and embedding images with HTML for resizing in Obsidian.
Ensure you have the following libraries installed before running the script:
pip install python-pptx transformers