| title | Paper Prism - Paper Image Extractor |
|---|---|
| emoji | 📑 |
| colorFrom | blue |
| colorTo | indigo |
| sdk | docker |
| app_port | 7860 |
| pinned | false |
| license | mit |
A premium, glassmorphism-inspired web tool designed for researchers to instantly extract high-quality figures and diagrams from academic papers using only a DOI or direct PDF upload.
Launch Paper Prism on Hugging Face Spaces
- DOI Smart Fetch: Paste a DOI, and we'll handle the rest via Sci-Hub mirrors/Unpaywall.
- Precision Extraction: Powered by
PyMuPDFto grab raw image bytes without quality loss. - Smart Filtering: Automatically hides tiny icons/logos and focuses on actual scientific figures.
- Social & Collaborative:
- Hall of Fame: Discover what researchers are trending globally.
- Live Multi-Country Hub: Real-time chat and leaderboard.
- Modern UX: Beautiful Neumorphic design with mobile-first responsiveness.
- SEO Optimized: Built to be discoverable by research communities worldwide.
- Backend: Python, FastAPI, APScheduler
- Extraction: PyMuPDF (fitz)
- Database/Real-time: Supabase (PostgreSQL & Realtime)
- Frontend: Vanilla JS (ES6+), CSS Grid/Flexbox, Neumorphism system
- Deployment: Docker on Hugging Face Spaces
The app includes a built-in Janitor Service that automatically:
- Truncates old chat logs to maintain performance.
- Rotates the Hall of Fame images based on popularity and age.
- Employs a keep-alive self-ping task to stay active.
This tool is created for educational and research purposes. Please ensure you comply with international copyright laws and journal access policies.
Created with ❤️ for the global research community.