Cross-Modal Generalizable Visual-Language Models from Inter-Modal Bidirectional Supervision for Accurate Pathology Image Recognition | doi.page