Reproduce “InfoVis: Unified Slide Image Generation and Editing” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Jiahao Wang, Xiyan Jiang, Chenyuan Wu, Haoqiang Zhang, Xin Luo, Yueze Wang, Defu Lian, Bo Zhang and 2 more
InfoVis trains a 9B image model for slide generation and editing across four tasks: generating from text, following a style reference, incorporating supplied visuals, and editing a slide. It uses presentation data and fuses trained task branches. On VisSlideBench, it scores 7.94, near Nano-Banana-2’s 7.99, and leads open-source models on all 11 metrics; editing is in-distribution and coverage is English-only.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.