arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2102.13558cs.CLcs.CV
收藏

自然语言视频本地化:基于Span的问答框架研究

Natural Language Video Localization: A Revisit in Span-based Question Answering Framework

Institute of High Performance Computing, A*STAR(高性能计算研究院(新加坡科技研究局));School of Computer Science and Engineering, Nanyang Technological University(南洋理工大学计算机科学与工程学院);Institute of Infocomm Research, A*STAR(信息通信研究院(新加坡科技研究局))

Hao Zhang, Aixin Sun, Wei Jing, Liangli Zhen, Joey Tianyi Zhou, Rick Siow Mong Goh

摘要:

自然语言视频定位(NLVL)的目的是从语义上对应于文本查询的未修剪视频中定位目标时刻。现有的方法主要从计算机视觉的角度解决NLVL问题,将其公式化为排名,锚或回归任务。这些方法在长视频上本地化时遭受大的性能下降。在这项工作中,我们从一个新的角度来解决NLVL,即,基于跨度的问答(QA),通过将输入视频视为文本段落。我们提出了一个视频跨度本地化网络(VSLNet),顶部的标准跨度为基础的QA框架(命名为VSLBase),以解决NLVL。VSLNet通过简单而有效的查询引导突出显示(QGH)策略解决了NLVL和基于跨度的QA之间的差异。QGH引导VSLNet在突出显示的区域内搜索匹配的视频范围。为了解决长视频的性能下降,我们进一步扩展VSLNet的VSLNet-L通过应用多尺度分割和拼接策略。VSLNet-L首先将未修剪的视频分割成短片段;然后,它预测哪个片段包含目标时刻并抑制其他片段的重要性。最后,剪辑片段连接,具有不同的置信度,以准确地定位目标时刻。在三个基准数据集上的大量实验表明,所提出的VSLNet和VSLNet-L优于最先进的方法; VSLNet-L解决了长视频性能下降的问题。我们的研究表明,基于跨度的QA框架是解决NLVL问题的有效策略。

英文摘要:

Natural Language Video Localization (NLVL) aims to locate a target moment from an untrimmed video that semantically corresponds to a text query. Existing approaches mainly solve the NLVL problem from the perspective of computer vision by formulating it as ranking, anchor, or regression tasks. These methods suffer from large performance degradation when localizing on long videos. In this work, we address the NLVL from a new perspective, i.e., span-based question answering (QA), by treating the input video as a text passage. We propose a video span localizing network (VSLNet), on top of the standard span-based QA framework (named VSLBase), to address NLVL. VSLNet tackles the differences between NLVL and span-based QA through a simple yet effective query-guided highlighting (QGH) strategy. QGH guides VSLNet to search for the matching video span within a highlighted region. To address the performance degradation on long videos, we further extend VSLNet to VSLNet-L by applying a multi-scale split-and-concatenation strategy. VSLNet-L first splits the untrimmed video into short clip segments; then, it predicts which clip segment contains the target moment and suppresses the importance of other segments. Finally, the clip segments are concatenated, with different confidences, to locate the target moment accurately. Extensive experiments on three benchmark datasets show that the proposed VSLNet and VSLNet-L outperform the state-of-the-art methods; VSLNet-L addresses the issue of performance degradation on long videos. Our study suggests that the span-based QA framework is an effective strategy to solve the NLVL problem.

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。

↑