One Figure, Every Canvas: Editable Flowchart Relayout via Agentic Pipeline
What it is
This paper proposes an agentic pipeline for aspect-ratio-adaptive flowchart relayout, designed to adapt ML pipeline figures to various canvas sizes without breaking connections or hallucinating content. It uses a three-stage process (Parse, Style, Layout) with critic agents to ensure structural fidelity and outputs editable draw.io XML.
The problem it solves
Existing methods for adapting flowcharts to different aspect ratios fail by stretching blocks, hallucinating content, or silently breaking edge connections, which misrepresents the underlying method.
What you could use it for
- Adapting ML paper figures for conference slides
- Generating social media teasers from technical diagrams
- Creating poster-ready flowcharts from standard paper layouts
Read the source ↗Back to the library →
Summary written by an AI model from the source; check the source before you rely on it.