PaperBanana

PaperBanana is an AI academic illustration generator that automates the creation of publication-ready scientific figures. It transforms paper text into methodology diagrams, statistical charts, and infographics using a closed-loop five-agent architecture. Designed specifically for researchers, it ensures figures are faithful, precise, and aesthetically polished, allowing scientists to focus on their core work.

How to Use

To use PaperBanana, simply enter a text description of the desired academic figure. Users can then select a visual style and let the AI generator create a publication-ready scientific illustration in seconds. PaperBanana supports various inputs, including descriptions for methodology diagrams, statistical charts, system architectures, and rough hand-drawn sketches for refinement.

Key Features

• AI generation of publication-ready methodology diagrams and system architectures
• Generates mathematically precise statistical plots using executable Python Matplotlib code
• Aesthetic enhancement and refinement of existing rough sketches or whiteboard notes
• Built on a closed-loop five-agent architecture for high fidelity and precision
• Creation of scientifically accurate educational infographics

Use Cases

• Generating model architectures, algorithm flows, and system pipeline illustrations for research papers (e.g., NeurIPS, ICML).
• Creating mathematically precise statistical plots from raw data input.
• Refining rough hand-drawn sketches into professional-grade figures suitable for publication.
• Producing educational infographics for supplementary materials or lecture slides.