arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2203.09306cs.CLcs.CV
收藏

在多模态BERT中发现结构知识

Finding Structural Knowledge in Multimodal-BERT

KU Leuven(鲁汶大学);University of Copenhagen(哥本哈根大学);Uppsala University(乌普萨拉大学)

Victor Milewski, Miryam de Lhoneux, Marie-Francine Moens

摘要:

在这项工作中,我们研究了在多模态BERT模型的嵌入中学习到的知识。更具体地说,我们探讨他们的能力存储的语法结构的语言数据和结构学习对象的视觉数据。为了达到这一目标,我们首先通过描述图像的句子的依赖性解析和图像中对象区域之间的依赖性,分别使语言和视觉效果的固有结构明确。我们称这种显式的可视化结构为\textit{scene tree},它基于语言描述的依赖关系树。广泛的探测实验表明,多模BERT模型不编码这些场景树。代码可在\url{https://github.com/VSJMilewski/multimodal-probes}。

英文摘要:

In this work, we investigate the knowledge learned in the embeddings of multimodal-BERT models. More specifically, we probe their capabilities of storing the grammatical structure of linguistic data and the structure learned over objects in visual data. To reach that goal, we first make the inherent structure of language and visuals explicit by a dependency parse of the sentences that describe the image and by the dependencies between the object regions in the image, respectively. We call this explicit visual structure the \textit{scene tree}, that is based on the dependency tree of the language description. Extensive probing experiments show that the multimodal-BERT models do not encode these scene trees.Code available at \url{https://github.com/VSJMilewski/multimodal-probes}.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑