Inspiration
As a college student drowning in textbooks, I kept stumbling across unintentionally funny images—like a judgmental owl in a biology book or awkward historical diagrams. I realized that books are full of meme-worthy content trapped in academic context. BookMeme was born from the idea: What if AI could dig through those pages and help us unleash the memes hidden in plain sight?
What it does
BookMeme turns any book into a meme generator. Upload a PDF or ePub, and it extracts all images, identifies the funniest ones using AI, and lets you caption them. It’s a meme factory powered by academia, built for fun.
How we built it
We used Python (Flask, Celery, PyMuPDF, OpenCV, HuggingFace Transformers) for the backend, and React.js for the frontend. It handles different book formats, extracts images, ranks them by meme potential, and offers a captioning interface. The system is modular for future features like social media integration and AI captioning.
Challenges we ran into
Due to limited time and headcount, we didn’t fully implement the AI-generated captions or social media integrations. Complex document structures and teaching AI to “understand” humor were also significant technical hurdles.
Accomplishments that we're proud of
We built a working prototype that extracts images from books and scores them for humor. The image grid, manual captioning tools, and meme formatting are all live. Even without full automation, it’s already fun and usable.
What's next for BookMeme
🤖 AI-generated captions 📲 Mobile version 🌐 One-click social sharing 🎨 Meme templates & Reddit-ready formatting 🖼️ Public meme gallery
Log in or sign up for Devpost to join the conversation.