Structure-aware Contrastive Learning for Diagram Understanding of Multimodal Models

VLMRe-assign community