Paper page - InfoVis: Unified Slide Image Generation and Editing

Source: https://huggingface.co/papers/2610.16784

Abstract

InfoVis is a unified 9B model for slide generation and editing trained on millions of native records and evaluated on a new multidimensional benchmark, achieving top open-source performance across text rendering, instruction following, visual grounding, design, and editing.

Generating and editing presentation slides requires models to combine accurate text rendering, instruction following, visual grounding, aesthetic design, and precise editing within one canvas. Existing work covers only subsets of these capabilities and lacks large-scale slide-native supervision and diagnostic evaluation. We formulate four slide-centric tasks and build a scalable data engine with over 13M continual-pretraining records and 1.9Msupervised fine-tuningrecords. We also introduceVisSlideBench, a benchmark of 1,527 manually screened instances that covers all four tasks with task-specific, multidimensional evaluation. Its automatic rankings agree with blinded, dimension-wise human rankings. Using these resources, we trainInfoVis, a unified 9Bslide generationand editing model. OnVisSlideBench,InfoVisleads every evaluated open-source model on all eleven metrics and scores 7.94 overall against 7.99 for Nano-Banana-2. Starting fromFLUX.2-klein-base-9B, slide-specific post-training adds 1.6 points overall, with the largest generation gains incontent faithfulness; the remaining gap to proprietary models is smallest incontent faithfulnessand largest inaesthetic quality. Built onInfoVis, our document-to-slide workflow achieves the second-highest Aesthetics score on the externalSlidesGen-Bench. We will release the model, code, data, and benchmark.

View arXiv pageView PDFProject pageGitHub2Add to collection

Get this paper in your agent:

hf papers read 2610\.16784

Don’t have the latest CLI?curl \-LsSf https://hf\.co/cli/install\.sh \| bash

Models citing this paper1

#### InfoVis/InfoVis Text-to-Image• 9B• Updated1 day ago • 26 • 5

Datasets citing this paper2

#### InfoVis/Data Viewer• Updated2 days ago • 9.9M • 956 • 2 #### InfoVis/VisSlideBench Viewer• Updated2 days ago • 1.53k • 47 • 2

Spaces citing this paper2

Collections including this paper2