TUhjnbcbe - 2021/12/27 0:32:00
本文梳理了个多模态相关的数据集,下面简要介绍各数据集的概况。1VisualQuestionAnswering(VQA)IntroducedbyAgrawaletal.inVQA:VisualQuestionAnsweringVisualQuestionAnswering(VQA)isadatasetcontainingopen-endedquestionsaboutimages.Thesequestionsrequireanunderstandingofvision,languageand