Stride improves infographic creation with scale-aware multiagent refinement
Stride: Scale-Aware Multi-Agent Evolutionary Refinement for Infographic Generation
Computer Vision and Pattern Recognition
Summary
Creating infographics is hard because the picture has to fit lots of visual and text rules all at once. The paper presents Stride, a system that fixes infographic images step-by-step by using two special agents: one that handles big-picture layout fixes and one that makes small, local changes. Stride only accepts improvements that pass a checker, which helps it get better with each try without messing up good parts. The authors show Stride works better than other similar systems on several infographic benchmarks.
What this means in practice
- •For graphic design teams: Automatically refine infographic layouts and visuals to comply with complex design constraints without manual reworking.
- •For marketing content creators: Generate high-quality infographics that meet branding and semantic guidelines through automated iterative improvement.
Authors
Yan Zhou, Yang Yang
Abstract
Infographic generation challenges text-to-image models because a single image must satisfy numerous visual, textual, and semantic constraints. Iterative agentic refinement with a multimodal verifier is a natural remedy, but existing methods face a scale mismatch: local edits cannot correct global composition errors, while full regeneration may destroy already-correct regions. We introduce Stride, a constraint-grounded multi-agent framework that treats refinement over any frozen generator as a history-aware evolution strategy. Stride separates constraint evaluation - whether each requirement is satisfied - from visual grounding - where and how it is violated. It then routes global revisions to a composer subagent and local corrections to a retoucher subagent. Each subagent conducts internal planning and generates multiple candidates, while Stride accepts a candidate only when it improves the verifier score, reducing regression across iterations. Because stochastic generators and imperfect verifiers make any single step unreliable, it is this combination of scale-aware routing, per-step redundancy, and monotonic acceptance that yields stable improvement across iterations. With fully open-weight generator-verifier systems, Stride outperforms proprietary systems including Nano-Banana-2.0 and Seedream-5.0-Pro on BizGenEval, matches GPT-Image-2.0 and surpasses Seedream-5.0-Pro on IGenBench, and achieves a new state-of-the-art score on WISE, while substantially outperforming existing agentic methods across all evaluated benchmarks.