arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2201.10654cs.CV
收藏

SA-VQA:面向视觉问答的视觉和语义表示的结构化对齐

SA-VQA: Structured Alignment of Visual and Semantic Representations for Visual Question Answering

Northwestern University(西北大学);Microsoft Research(微软研究院)

Peixi Xiong, Quanzeng You, Pei Yu, Zicheng Liu, Ying Wu

摘要:

可视化问答(Visual Question Questioning,VQA)是一个备受业界和学术界关注的问题。作为一个多模态任务,它是具有挑战性的,因为它不仅需要视觉和文本的理解,但也有能力调整跨模态表示。以前的方法广泛采用实体级的对齐,如视觉区域和它们的语义标签之间的相关性,或跨问题词和对象特征的相互作用。这些尝试旨在改善跨通道表征,而忽略了它们之间的内在联系。相反,我们建议应用结构化对齐,它与视觉和文本内容的图形表示一起工作,旨在捕捉视觉和文本模态之间的深层联系。然而,它是不平凡的表示和整合结构化的比对图。在这项工作中,我们试图解决这个问题,首先将不同的模态实体转换成连续的节点和邻接图,然后将它们结构化对齐。正如我们的实验结果所示,这样的结构化对齐提高了推理性能。此外,我们的模型还表现出更好的可解释性为每个生成的答案。该模型在没有任何预训练的情况下,在GQA数据集上的性能优于最先进的方法,并在VQA-v2数据集上击败了非预训练的最先进的方法。

英文摘要:

Visual Question Answering (VQA) attracts much attention from both industry and academia. As a multi-modality task, it is challenging since it requires not only visual and textual understanding, but also the ability to align cross-modality representations. Previous approaches extensively employ entity-level alignments, such as the correlations between the visual regions and their semantic labels, or the interactions across question words and object features. These attempts aim to improve the cross-modality representations, while ignoring their internal relations. Instead, we propose to apply structured alignments, which work with graph representation of visual and textual content, aiming to capture the deep connections between the visual and textual modalities. Nevertheless, it is nontrivial to represent and integrate graphs for structured alignments. In this work, we attempt to solve this issue by first converting different modality entities into sequential nodes and the adjacency graph, then incorporating them for structured alignments. As demonstrated in our experimental results, such a structured alignment improves reasoning performance. In addition, our model also exhibits better interpretability for each generated answer. The proposed model, without any pretraining, outperforms the state-of-the-art methods on GQA dataset, and beats the non-pretrained state-of-the-art methods on VQA-v2 dataset.

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。

↑