Knossos and Ariadne: Benchmarking and Learning Complete Diagram Topology Extraction with Vision-Language Models

Excerpt

arXiv:2610.04721v1 Announce Type: cross Abstract: Structural diagrams are widely used to represent complex systems and relational information across scientific, engineering, procedural, and spatial domains. Recent vision-language models (VLMs) have become increasingly capable of recognizing diagram elements and reasoning about their content, while complete diagram topology extraction remains comparatively underexplored. In this paper, we study diagram-to-graph topology extraction: extracting all