AutoFigure: Generating and Refining Publication-Ready Scientific Illustrations

By: Minjun Zhu, Zhen Lin, Yixuan Weng, Panzhong Lu, Qiujie Xie, Yifan Wei, Sifan Liu, Qiyao Sun, Yue Zhang

Published: 2026-02-04

View on arXiv →
#cs.AI

Abstract

High-quality scientific illustrations are crucial for effectively communicating complex scientific and technical concepts, yet their manual creation remains a well-recognized bottleneck in both research and publishing. We present FigureBench, the first large-scale benchmark for generating scientific illustrations from long-form scientific texts. Moreover, we propose AutoFigure, the first agentic framework that automatically generates high-quality scientific illustrations based on long-form scientific text. Specifically, before rendering the final result, AutoFigure engages in extensive thinking, recombination, and validation to produce a layout that is both structurally sound and aesthetically refined, outputting a scientific illustration that achieves both structural completeness and aesthetic appeal. The results demonstrate that AutoFigure consistently surpasses all baseline methods, producing publication-ready scientific illustrations.

FEEDBACK

Projects

No projects yet