BriefGPT.xyz
Feb, 2023
利用联合自监督学习的医学视觉问答
Medical visual question answering using joint self-supervised learning
HTML
PDF
Yuan Zhou, Jing Mei, Yiqin Yu, Tanveer Syeda-Mahmood
TL;DR
本研究提出一种编码器-解码器框架,利用自注意机制跨图像文本双模态表示,并通过自监督多任务学习在大规模医学图像字幕数据集上进行预训练,并在小规模医学VQA数据集上进行微调,取得了比基线和SOTA方法更好的性能。
Abstract
visual question answering
(VQA) becomes one of the most active research problems in the
medical imaging
domain. A well-known VQA challenge is the intrinsic diversity between the image and text modalities, and in
→