Human Pathology Image-Text Dataset
收藏arXiv2025-09-30 收录
下载链接:
https://github.com/ddw2AIGROUP2CQUPT/PA-LLaVA
下载链接
链接失效反馈官方服务:
资源简介:
该数据集是一个针对病理学领域构建的特定领域数据集,它通过清洗公开的医学图像-文本数据,以便在病理学中进行对齐。此外,该数据集专门为计算病理学领域量身定制,用于训练专门的视觉编码器,其任务是进行视觉问题回答(Visual Question Answering, VQA)。
This is a domain-specific dataset constructed for the field of pathology. It cleans publicly available medical image-text pairs to enable cross-modal alignment in the context of pathology. Moreover, this dataset is specifically tailored for the computational pathology domain, and is designed to train specialized visual encoders for Visual Question Answering (VQA) tasks.
提供机构:
Authors of the paper



