Work

Global Contrastive Learning with High-Quality Data in Large Vision-Language Models for Pathological Question Answering